[ { "productId": "anrok", "storyId": "agent-reads-tax-state", "verdict": "partial", "quality": 4, "confidence": "medium", "rationale": "Anrok's API exposes some relevant read/act endpoints (filings, tax-ID/registration lists, address resolution, transaction create/void/refund) and marketing for 'headless-tax' explicitly says you can 'hand it to your agents', plus AI Atlas offers grounded Q&A over transactions/registrations/filings/certificates. However there is no evidence of an MCP surface, no mention of scoped/limited API credentials, no explicit nexus-status or accrued-liability endpoint, and probes show llms.txt and openapi discovery endpoints 404ing, undercutting easy machine/agent onboarding. Missing for 10: MCP server or endpoint, scoped-credential/OAuth support, explicit nexus-status and accrued-liability read endpoints, working machine-readable API discovery (llms.txt/openapi).", "evidenceIds": [ "anrok-docs-10", "anrok-docs-12", "anrok-docs-15", "anrok-docs-16", "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "agentic-agent-docs", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "A direct probe for llms.txt returned 404, and no agent-oriented docs endpoint was found; OpenAPI probes also failed. No evidence supports pointing an agent at machine-readable docs.", "evidenceIds": [ "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "agentic-ai-insights", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "Anrok's AI Atlas feature explicitly lets users ask natural-language questions about their sales tax and receive AI-generated answers with reasoning, grounded in their own transactions, registrations, filings, and certificates. This directly matches the story of getting AI-generated insights from within the product's own data. Missing for 10: independent/hands-on corroboration of AI Atlas quality, and more detail on the range/depth of insights or suggestions beyond Q&A.", "evidenceIds": [ "anrok-docs-15", "anrok-docs-16" ] }, { "productId": "anrok", "storyId": "agentic-autonomous-automation", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Anrok's marketing mentions background automations like automatic jurisdiction registration completion, early tax-obligation alerts, and a headless API that can be 'handed to your agents,' implying some autonomous background processing (tax calc, filings, registrations). However there's no evidence of a user-configurable automation builder or explicit agentic scheduling system that an AI-native user could set up themselves. Missing for 10: documentation of a user-facing automation/workflow builder, evidence of agent-triggered autonomous tasks beyond marketing claims, and independent confirmation these run without manual oversight.", "evidenceIds": [ "anrok-docs-16", "anrok-docs-18", "anrok-docs-20", "anrok-docs-15" ] }, { "productId": "anrok", "storyId": "agentic-builtin-assistant", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Anrok's 'Atlas' feature is described as a built-in AI assistant that lets users 'ask it anything about your sales tax' with reasoning grounded in the customer's own transactions, registrations, filings, and certificates, and the 'headless-tax' page mentions handing tax work 'to your agents.' However, evidence is limited to marketing copy about answering questions and general agent-friendliness, without documentation of the assistant executing multi-step tasks (e.g., filing, remediation) autonomously. Missing for 10: concrete examples of the assistant taking action (not just Q&A), technical/architecture docs on the assistant's task-execution scope, and independent/hands-on verification of its capabilities.", "evidenceIds": [ "anrok-docs-15", "anrok-docs-16" ] }, { "productId": "anrok", "storyId": "agentic-headless", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "Anrok is explicitly headless/API-first ('Run tax through Anrok's APIs, or hand it to your agents') with a full REST API covering customers, transactions, certificates, filings, and tax calculation that can be scripted for automation/CI pipelines. Missing for 10: explicit CI/CD pipeline examples, official SDKs/CLI tooling, and a working machine-readable openapi.json/llms.txt (probes show 404s) that would make automated integration turnkey.", "evidenceIds": [ "anrok-docs-16", "anrok-docs-2", "anrok-docs-3", "anrok-docs-6", "anrok-docs-4", "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "agentic-mcp-client", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Anrok is a tax compliance SaaS product, not an AI agent client; the ability to plug MCP servers into it as tools is a client-role capability that doesn't apply to this product category. No evidence suggests Anrok itself functions as an MCP client.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "agentic-mcp-server", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Anrok exposes a REST API and mentions 'hand it to your agents' but there is no evidence of an official MCP server for connecting AI agents; llms.txt and openapi probes both 404.", "evidenceIds": [ "anrok-docs-16", "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "agentic-nl-commands", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Anrok's AI Atlas feature lets users 'ask it anything about your sales tax' and get grounded answers, showing a natural-language query interface, and headless-tax mentions handing tax operations to agents. However, evidence only shows NL-based Q&A/insight retrieval, not full operational control (e.g., creating transactions, managing filings) via natural language commands. missing for 10: evidence of NL commands executing actions (not just queries), documentation of command syntax/scope, independent verification of AI Atlas performance.", "evidenceIds": [ "anrok-docs-15", "anrok-docs-16" ] }, { "productId": "anrok", "storyId": "agentic-official-cli", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Anrok offers an API and headless tax product, but no evidence of an official CLI tool; probes for llms.txt and openapi endpoints both returned 404s, and no docs mention a CLI.", "evidenceIds": [ "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "agentic-public-api", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Anrok exposes a well-documented public API covering customers, certificates, tax calculation, transactions, filings, and product mappings, with a dedicated tutorials/integration overview and OpenAPI spec referenced, plus explicit messaging that agents can be handed the API (\"Run tax through Anrok's APIs, or hand it to your agents\"). Missing for 10: independent third-party corroboration of API usage and confirmation of a discoverable machine-readable spec (probe found openapi.json/llms.txt 404s despite an openapi.yaml reference).", "evidenceIds": [ "anrok-docs-2", "anrok-docs-3", "anrok-docs-5", "anrok-docs-6", "anrok-docs-10", "anrok-docs-11", "anrok-docs-12", "anrok-docs-13", "anrok-docs-16", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "agentic-scoped-keys", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of API key scoping, permission levels, or credential issuance mechanisms for agents; docs describe tax API endpoints but nothing about least-privilege or scoped credential issuance.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "agentic-sdks", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows Anrok has a REST API with OpenAPI schema references, but there is no mention of official client SDKs (e.g., Python, Node, Java libraries) for developers to build against; probes even show openapi.json/llms.txt discovery paths returning 404, suggesting no packaged SDK or machine-readable spec artifact is publicly distributed. missing for 10: named official SDK packages, language-specific client libraries, SDK installation/quickstart docs.", "evidenceIds": [ "anrok-probe-1", "anrok-probe-2", "anrok-docs-4" ] }, { "productId": "anrok", "storyId": "agentic-webhooks", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of any webhook subscription mechanism; all documented capabilities are REST endpoints for customers, transactions, filings, and certificates, with no mention of webhooks or event subscriptions.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "ai-taxability-answers", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "Anrok's 'AI Atlas' is explicitly described as answering sales-tax questions 'with the reasoning shown, grounded in your own transactions, registrations, filings, and certificates' — a purpose-built tax AI over its own rule/data base rather than a generic chatbot. Missing for 10: independent/hands-on validation of answer accuracy, detail on invoice-level 'why did this tax this way' explanations, and citation format specifics.", "evidenceIds": [ "anrok-docs-15", "anrok-docs-16" ] }, { "productId": "anrok", "storyId": "api-interactive-docs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows Anrok's API docs describe endpoints (customers, tax IDs, address resolution, etc.) and reference an openapi.yaml spec, but nothing indicates an interactive reference with runnable/try-it examples; probes even show that common interactive-doc artifacts (openapi.json, swagger.json, llms.txt) return 404. Missing for 10: any mention of a 'try it' console, code sandbox, or runnable request builder in the docs.", "evidenceIds": [ "anrok-docs-2", "anrok-docs-3", "anrok-docs-4", "anrok-docs-10", "anrok-docs-11", "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "api-machine-spec", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Anrok's docs repeatedly reference a live openapi.yaml file (e.g., anrok-docs-4, anrok-docs-10-13) suggesting a machine-readable OpenAPI spec exists, but probes for common JSON/well-known OpenAPI paths and llms.txt all 404, and no direct confirmation of a downloadable spec file was found. Missing for 10: a verified working download link for openapi.yaml, an llms.txt or AI-friendly spec index, and independent confirmation the spec is complete/machine-consumable.", "evidenceIds": [ "anrok-docs-4", "anrok-docs-10", "anrok-docs-11", "anrok-docs-12", "anrok-docs-13", "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "api-sandbox", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence pack item mentions a sandbox, test mode, or non-production environment for testing API integrations; the createEphemeral endpoint is described as a tax estimation tool, not a sandbox for isolating test data from production. Missing for 10: any mention of sandbox/test environment, test API keys, or documented separation between test and live data.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "api-versioning-policy", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence of API versioning scheme or a documented deprecation policy anywhere in the docs pack; probes even show standard OpenAPI/discovery files returning 404. missing for 10: versioning scheme documentation, explicit deprecation policy, changelog or migration guides.", "evidenceIds": [ "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "autofiling-remittance", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Evidence shows Anrok can fetch filings and supports automatic registration across jurisdictions, implying some filing automation, but there is no mention of remittance, on-time guarantee, or penalty coverage anywhere in the pack. missing for 10: explicit description of end-to-end return preparation/remittance workflow, guarantee language covering late-filing penalties, evidence of on-time filing across all registered jurisdictions.", "evidenceIds": [ "anrok-docs-12", "anrok-docs-18" ] }, { "productId": "anrok", "storyId": "automation-bulk-operations", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Anrok's API supports listing customers and creating many-to-1 product mappings across a catalog, which implies some bulk-style operations, but there is no explicit documentation of batch create/update/delete endpoints or bulk transaction processing across many items at once. missing for 10: dedicated bulk/batch API endpoints, evidence of bulk certificate or transaction processing, and independent confirmation of scale handling.", "evidenceIds": [ "anrok-docs-3", "anrok-docs-13", "anrok-docs-14" ] }, { "productId": "anrok", "storyId": "automation-rules-engine", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Anrok's evidence covers tax calculation APIs, certificate/registration tracking, and an AI Q&A assistant (AI Atlas), but nothing describes a user-defined rules engine that automatically triggers actions on events (e.g., 'if nexus threshold crossed, auto-register' as a configurable rule). 'Early alerts' and 'automatic registration completion' are pre-built product features, not user-definable trigger/action rules.", "evidenceIds": [ "anrok-docs-15", "anrok-docs-16", "anrok-docs-20" ] }, { "productId": "anrok", "storyId": "automation-scheduled-jobs", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Anrok is a tax compliance/API platform; no evidence describes any scheduling of recurring jobs or workflows (cron-like automation, scheduled syncs, recurring filings automation via API). Filings, registrations, and calculations are described as on-demand API calls or automatic processes, not user-schedulable recurring jobs.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "automation-versioned-workflows", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Anrok is a sales tax calculation/compliance product, not an automation-building platform; there is no concept of user-defined automations to version, review, or roll back. This story applies to workflow/agent-builder tools, not a tax engine API.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "billing-stack-integrations", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Anrok documents a Shopify integration (verifying Shopify's own tax calc) and a general API for transactions, customers, and product mappings, but the mapping endpoint (many-to-1 product mapping) and createOrUpdate/negation flows show developers still must build the plumbing to sync billing-system data rather than plugging into a pre-built connector library covering multiple billing/payment/commerce platforms. Missing for 10: a documented catalog of maintained native integrations (e.g., Stripe, Chargebee, Recurly, other commerce platforms) beyond Shopify, and evidence that these connectors sync automatically without developer-built API glue.", "evidenceIds": [ "anrok-docs-19", "anrok-docs-13", "anrok-docs-6", "anrok-docs-8", "anrok-docs-9", "anrok-docs-16" ] }, { "productId": "anrok", "storyId": "calculation-accuracy-guarantee", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence pack item mentions an accuracy guarantee, indemnification, warranty, or audit-defense/support commitment tied to Anrok's tax calculations; all citations cover API mechanics, product features, and dashboards. Absence of any documented guarantee for this applicable finance/tax axis yields 'none'.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "compliant-invoice-fields", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "Anrok computes tax amounts, resolves jurisdictions/addresses, and collects/validates customer tax IDs per jurisdiction (docs-10, docs-11), which supports registration-number and locale-rule inputs into billing systems, but there is no evidence it generates or validates full compliant invoice documents (e.g., specific required VAT invoice fields, reverse-charge notices, locale formatting) itself. missing for 10: evidence of actual invoice-field/template compliance output, confirmation that generated invoices meet locale-specific legal requirements, and any hands-on/independent verification of this.", "evidenceIds": [ "anrok-docs-10", "anrok-docs-11", "anrok-docs-17", "anrok-docs-18" ] }, { "productId": "anrok", "storyId": "exempt-customer-api", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Docs confirm certificates (which typically carry exemption reasons) can be uploaded via API alongside the calculation/transaction/return endpoints, implying exemption data could flow into the tax engine, but there is no explicit documentation of an API field to mark a customer/transaction exempt with a reason, nor confirmation that this reason is surfaced in reporting or on the filed return. Missing for 10: explicit API parameter/endpoint for setting exemption status+reason, and evidence that the reason appears in reports and filed returns.", "evidenceIds": [ "anrok-docs-4", "anrok-docs-21", "anrok-docs-6", "anrok-docs-12" ] }, { "productId": "anrok", "storyId": "exemption-certificate-management", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Anrok supports certificate collection and storage via UI/API, and pricing page states certificates are \"collected, verified, and stored,\" with customer search enabling pre-billing certificate upload. However, there is no evidence of expiry tracking or explicit automatic application of certificates at tax-calculation time within the evidence pack. missing for 10: expiry tracking/alerting on certificates, explicit description of certificates being auto-applied during tax calculation, independent verification of the certificate workflow.", "evidenceIds": [ "anrok-docs-1", "anrok-docs-4", "anrok-docs-21" ] }, { "productId": "anrok", "storyId": "filing-calendar-visibility", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Evidence only shows a generic API endpoint to 'fetch filings' from Anrok, implying some filing-tracking capability, but there is no documentation of a calendar of upcoming returns, per-filing status detail, remitted amounts, or downloadable confirmations for submitted returns. Missing for 10: calendar view of upcoming filings, per-filing status detail, remitted amount reporting, downloadable filing confirmations.", "evidenceIds": [ "anrok-docs-12" ] }, { "productId": "anrok", "storyId": "global-vat-gst", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Evidence shows global exposure tracking (world exposure map, UK registration threshold tracker, tax obligations tracked across 13 countries, automatic registration in 13 jurisdictions) and filings API support, indicating international VAT/GST coverage beyond US sales tax. However, there's no explicit mention of EU OSS/IOSS schemes, Canada or Australia specifically, or detail on how foreign registration thresholds are tracked/filed beyond the UK example and generic '13 jurisdictions.' Missing for 10: explicit EU OSS/IOSS support, named coverage of Canada/Australia, and detailed foreign filing mechanics beyond dashboard alerts.", "evidenceIds": [ "anrok-docs-17", "anrok-docs-18", "anrok-docs-12", "anrok-docs-20" ] }, { "productId": "anrok", "storyId": "historical-exposure-analysis", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows Anrok tracks nexus/exposure prospectively (world exposure map, registration threshold tracker, early alerts) but nothing describes importing historical transactions to backdate exposure or calculate estimated liability, penalties, and interest per jurisdiction. missing for 10: historical transaction import/backfill feature, backdated exposure calculation, penalty and interest estimation per jurisdiction.", "evidenceIds": [ "anrok-docs-17", "anrok-docs-20" ] }, { "productId": "anrok", "storyId": "liability-reporting", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Anrok's docs show jurisdiction-level exposure dashboards and a filings-fetch API, implying some accrued/owed visibility and filing-ready data, but there is no explicit evidence of a collected-vs-owed liability view, GL-formatted exports, or accountant-consumable outputs. missing for 10: explicit collected vs owed liability breakdown, GL/accounting export formats, accountant-facing report examples.", "evidenceIds": [ "anrok-docs-12", "anrok-docs-17", "anrok-docs-18", "anrok-docs-20" ] }, { "productId": "anrok", "storyId": "marketplace-facilitator-handling", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence pack items mention marketplace facilitator transactions, facilitator-collected tax tracking, or logic to avoid double-remittance on marketplace sales — the docs cover general transaction/refund/certificate APIs but nothing specific to marketplace facilitator nexus handling.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "nexus-threshold-monitoring", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Anrok's marketing docs explicitly claim 'Early alerts before you owe anywhere' and show a world exposure map with a UK registration threshold tracker, indicating nexus monitoring across jurisdictions with proactive alerting. However, there is no detailed documentation of how thresholds are tracked per state/country, alert timing/mechanics, or transaction-count-based nexus logic. missing for 10: technical detail on threshold-tracking logic per state/country, transaction-count thresholds (not just revenue), alert delivery mechanics/timing, and independent/hands-on verification that alerts fire before crossing thresholds.", "evidenceIds": [ "anrok-docs-17", "anrok-docs-20", "anrok-docs-18" ] }, { "productId": "anrok", "storyId": "omnichannel-import", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Anrok's API clearly supports transaction ingestion (createOrUpdate, void, refund/negation endpoints) and product mapping across billing systems (anrok-docs-6,7,8,9,13), and it explicitly recalculates/verifies Shopify tax data (anrok-docs-19), showing at least one storefront connector beyond the core API. However, evidence does not show a broader library of connectors to other processors/storefronts (e.g., Stripe, other marketplaces) or explicit historical-data backfill tooling. missing for 10: documented connectors to multiple processors/storefronts beyond Shopify, explicit bulk/historical backfill API or import tool, independent confirmation of multi-source coverage.", "evidenceIds": [ "anrok-docs-6", "anrok-docs-7", "anrok-docs-8", "anrok-docs-9", "anrok-docs-13", "anrok-docs-19" ] }, { "productId": "anrok", "storyId": "openness-api-parity", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Anrok's API documentation covers most core operations (customers, certificates, transactions, tax IDs, address resolution, filings, product mappings) and it markets itself as 'headless tax' with 'one engine' powering both UI and API/agent workflows, suggesting strong parity. However, dashboard-style features (world exposure map, registration threshold tracker, AI Atlas natural-language assistant) are only described as UI experiences with no documented API equivalent, and probes show no discoverable OpenAPI spec or llms.txt for full API surface verification. Missing for 10: explicit statement/documentation of full UI-API parity, API access to dashboard/exposure-map and AI Atlas features, and a verifiable machine-readable API spec.", "evidenceIds": [ "anrok-docs-4", "anrok-docs-12", "anrok-docs-16", "anrok-docs-17", "anrok-docs-15", "anrok-probe-1", "anrok-probe-2" ] }, { "productId": "anrok", "storyId": "openness-full-export", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "There is no evidence of a bulk data export feature or open-format data portability (e.g., CSV/JSON export of all transactions, certificates, filings) for users to leave with their data; API endpoints only support fetching individual records via calls, not a full export mechanism, and probes for llms.txt/openapi.json artifacts returned 404s.", "evidenceIds": [ "anrok-probe-1", "anrok-probe-2", "anrok-docs-2", "anrok-docs-3" ] }, { "productId": "anrok", "storyId": "openness-open-license", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Anrok is a proprietary SaaS tax platform, not open-source software; source code availability under an open license is not a relevant axis for this product category.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "openness-self-host", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Anrok is a SaaS tax compliance platform, not open-source infrastructure software; self-hosting is a category error for this kind of hosted API service.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "privacy-data-residency", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence in the pack addresses data residency, regional storage options, or data location controls for Anrok's platform; all evidence concerns tax calculation, filings, and API endpoints unrelated to data residency.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "privacy-no-training", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence in the pack addresses AI training data usage or opt-out policies for Anrok's AI features (e.g., AI Atlas); this is a plausible axis for a product with AI-driven features, but nothing documents a data-training opt-out.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "privacy-retention-controls", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers Anrok's tax calculation, filing, and certificate APIs but contains no mention of data retention policies, data deletion controls, or privacy/data lifecycle management features for AI-native users. No documentation addresses how users can control or delete stored data.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "privacy-telemetry-optout", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Anrok is a tax compliance SaaS platform, not an AI agent/tool whose telemetry or usage-tracking opt-out would be a meaningful buyer axis; the evidence pack contains no relevant material either way.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "product-tax-codes", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Anrok's API supports mapping catalog products to tax categories via product IDs and many-to-1 mappings (each unique tax category needs only one product ID), and jurisdiction-aware tax calculation/address resolution is documented. However, there's no explicit documentation showing the specific vendor tax category taxonomy (SaaS, digital goods, services, physical products) or a UI/API walkthrough of assigning each of those category types across jurisdictions. missing for 10: explicit list/definition of tax category taxonomy (SaaS vs digital goods vs services vs physical goods), concrete UI/API example mapping catalog items to these categories, independent corroboration of category-based jurisdictional taxation accuracy.", "evidenceIds": [ "anrok-docs-13", "anrok-docs-14", "anrok-docs-11", "anrok-docs-10" ] }, { "productId": "anrok", "storyId": "rate-rule-currency", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "The evidence pack covers API endpoints, certificates, filings, and dashboards, but contains no claims about automatic rate/rule updates across jurisdictions, no changelog or effective-date documentation, and no statement that customers never need to redeploy for rate changes. This is a fair and central axis for a tax-calculation vendor, but no supporting evidence exists.", "evidenceIds": [] }, { "productId": "anrok", "storyId": "realtime-tax-calculation", "verdict": "partial", "quality": 7, "confidence": "medium", "rationale": "Anrok's API docs describe a createEphemeral endpoint for real-time tax estimation during checkout and a dedicated Address Resolution endpoint that normalizes a customer address to determine taxable jurisdiction(s), supporting rooftop-level (not ZIP-only) accuracy. However, there is no explicit performance/latency benchmark or independent hands-on confirmation that calls are fast enough for checkout-critical paths. Missing for 10: documented latency/SLA numbers for checkout use, independent/third-party verification of real-time performance, and explicit comparison against ZIP-code-level competitors.", "evidenceIds": [ "anrok-docs-5", "anrok-docs-11", "anrok-docs-6" ] }, { "productId": "anrok", "storyId": "refund-adjustment-handling", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Anrok's API docs explicitly cover refund/credit-note workflows: createOrUpdate for negative transactions, createNegation for full negation plus a new corrected transaction for partial refunds, and each saved transaction is tied to the appropriate tax return, implying downstream liability and filing effects. However, there's no explicit evidence describing how these adjustments flow through reporting dashboards or that manual journal fixes are entirely eliminated across all reporting/filing steps. Missing for 10: explicit documentation or independent confirmation that negative amounts automatically reconcile in liability reports and filings without any manual reconciliation, and real-world/hands-on validation of this flow.", "evidenceIds": [ "anrok-docs-6", "anrok-docs-8", "anrok-docs-9", "anrok-docs-7" ] }, { "productId": "anrok", "storyId": "saas-digital-taxability", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Anrok's product-category mapping (docs-13, docs-14) shows it classifies products into tax categories, and AI Atlas (docs-15) lets users 'ask anything about sales tax' with reasoning shown, grounded in transactions/registrations/filings/certificates — suggesting visible research behind determinations. However, there's no direct evidence addressing the specific hard cases named (state-by-state SaaS taxability differences, digital-goods nuances, or bundled-offering treatment) or how the reasoning trail covers those scenarios. Missing for 10: explicit documentation of SaaS taxability variance by state, digital-goods rule handling, bundled-offering tax logic, and independent verification that Atlas's 'reasoning shown' actually resolves these edge cases correctly.", "evidenceIds": [ "anrok-docs-13", "anrok-docs-14", "anrok-docs-15", "anrok-docs-16" ] }, { "productId": "anrok", "storyId": "state-registrations", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Anrok's marketing claims 'automatic registration completion in 13 jurisdictions' and an exposure/threshold tracker, suggesting some in-product registration handling, but there is no evidence of per-registration pricing transparency or a dedicated status-tracking workflow for permit filings. missing for 10: per-registration pricing disclosure, explicit status tracking UI/API for registrations, confirmation of full state/local filing scope.", "evidenceIds": [ "anrok-docs-17", "anrok-docs-18", "anrok-docs-20" ] }, { "productId": "anrok", "storyId": "transaction-audit-trail", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Anrok's API retains transaction data (createOrUpdate/createNegation saved and tied to filings), address resolution is a distinct queryable endpoint, and product category mapping is stored per tax category; AI Atlas explicitly says answers are grounded in transactions, registrations, filings, and certificates, implying underlying data is retrievable. However, there's no explicit documentation of a per-transaction audit trail showing the exact rate and rule set applied, nor a dedicated export/audit report feature. Missing for 10: explicit rate/rule audit trail per transaction, a documented export function for audit purposes, and independent confirmation of reconstructability beyond docs claims.", "evidenceIds": [ "anrok-docs-6", "anrok-docs-11", "anrok-docs-12", "anrok-docs-13", "anrok-docs-14", "anrok-docs-15" ] }, { "productId": "anrok", "storyId": "vat-id-validation", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows Anrok collects and validates customer tax IDs for applicable jurisdictions (anrok-docs-10) and tracks global VAT obligations (anrok-docs-17, anrok-docs-18), but there is no evidence of real-time validation against official government registries (e.g., VIES) or of reverse-charge mechanics being automatically applied to B2B cross-border transactions.", "evidenceIds": [ "anrok-docs-10", "anrok-docs-17", "anrok-docs-18" ] }, { "productId": "anrok", "storyId": "vda-remediation", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Evidence shows Anrok tracks exposure, thresholds, and sends early alerts before liability accrues, but nothing in the pack mentions voluntary disclosure agreements, back-filing support, or any remediation path for past-due/uncollected tax periods.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "agent-reads-tax-state", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Avalara ships an official MCP server exposing nexus definitions, transaction retrieval, tax code recommendations, and general compliance data (avalara-docs-6,7,8,13; avalara-probe-3), plus REST APIs/SDKs for creating nexus and transactions (avalara-docs-4,5,3) — a real agent-accessible surface. However the evidence never shows endpoints for 'accrued liability' summaries or 'upcoming filings' schedules, nor details on scoped/least-privilege credential issuance for agents. Missing for 10: accrued liability reporting endpoint, filing calendar/upcoming filings API or MCP tool, explicit scoped-credential/auth-scoping documentation for agent use.", "evidenceIds": [ "avalara-docs-6", "avalara-docs-7", "avalara-docs-8", "avalara-docs-13", "avalara-docs-5", "avalara-docs-4", "avalara-probe-3" ] }, { "productId": "avalara", "storyId": "agentic-agent-docs", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Direct probe confirms llms.txt returns HTTP 200 at developer.avalara.com/llms.txt, and docs reference agent-oriented .md pages (e.g. avatax.md, mcp.md) plus a dedicated mcp-servers.md, showing an AI-agent-friendly documentation surface. Missing for 10: independent/community confirmation of an agent successfully consuming llms.txt in practice, and no OpenAPI machine-readable spec found (404s on probe).", "evidenceIds": [ "avalara-probe-1", "avalara-docs-6", "avalara-probe-3", "avalara-probe-2" ] }, { "productId": "avalara", "storyId": "agentic-ai-insights", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Docs mention AI-powered tax code recommendations with confidence levels (avalara-docs-7) and conversational-query analytics for spotting transaction patterns/anomalies (avalara-docs-12), which are AI-generated insight/suggestion features embedded in the product. However, these are narrow (tax-code classification, basic query analytics) rather than broad, proactive insight generation across the product's data, and there's no independent/hands-on evidence validating quality or scope. Missing for 10: evidence of broader analytical dashboards or proactive anomaly/insight surfacing beyond tax-code suggestions, and independent corroboration of the conversational analytics feature's real-world performance.", "evidenceIds": [ "avalara-docs-7", "avalara-docs-12", "avalara-docs-6" ] }, { "productId": "avalara", "storyId": "agentic-autonomous-automation", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "AvaTax's evidence shows API/MCP integration for on-demand tax calculation and lookups (transactions, nexus, tax-code suggestions), but nothing describes a scheduling, workflow, or trigger mechanism that lets an AI-native user configure automations that run autonomously in the background without a calling system initiating each request. missing for 10: evidence of scheduled/triggered background jobs, webhooks, or an autonomous workflow engine.", "evidenceIds": [ "avalara-docs-6", "avalara-docs-7", "avalara-docs-8", "avalara-docs-13" ] }, { "productId": "avalara", "storyId": "agentic-builtin-assistant", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Evidence only hints at AI features (AI-powered tax code recommendations, and 'conversational queries' for analyzing accounts) rather than a documented built-in assistant a user can delegate broader tasks to; the MCP server (docs-6/7/8/13, probe-3) is framed as an API surface for external AI agents to call into AvaTax, not an assistant embedded in the product for the user. Missing for 10: a documented in-product chat/assistant UI, description of delegable task scope, and evidence of user-facing conversational workflows beyond a single marketing line.", "evidenceIds": [ "avalara-docs-12", "avalara-docs-7", "avalara-docs-6", "avalara-probe-3" ] }, { "productId": "avalara", "storyId": "agentic-headless", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "AvaTax exposes REST v2 APIs and SDKs in multiple languages (C#, Java, JS, Python, PHP, Ruby) that can be scripted headlessly for CI/automation, and API endpoints like CreateTransaction/CreateNexus are clearly designed for programmatic, non-interactive use. However, there is no explicit CI/automation documentation, no CLI tool, and no discussion of authentication patterns, rate limits, or idempotency for automated pipelines. missing for 10: explicit CI/headless usage docs or examples, a dedicated CLI, and evidence of automation-friendly auth/rate-limit handling for unattended pipelines.", "evidenceIds": [ "avalara-docs-3", "avalara-docs-4", "avalara-docs-5", "avalara-docs-9" ] }, { "productId": "avalara", "storyId": "agentic-mcp-client", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "AvaTax is a tax-calculation API/SaaS product, not an AI agent or client application that would itself consume MCP servers' tools; the evidence only shows Avalara exposing its own MCP server (docs-6,7,8,13, probe-3) for others to call into, which is the opposite direction from this story's client-side 'plug MCP servers into this product' capability. This axis does not apply to a service of this kind.", "evidenceIds": [ "avalara-docs-6", "avalara-docs-7", "avalara-docs-8", "avalara-docs-13", "avalara-probe-3" ] }, { "productId": "avalara", "storyId": "agentic-mcp-server", "verdict": "full", "quality": 8, "confidence": "medium", "rationale": "Avalara documents an official MCP server (mcp.md, mcp-servers.md) exposing AvaTax capabilities—tax calculation, transaction management, nexus handling, AI-powered tax code recommendations—directly usable by AI agents, and a probe confirms the server documentation is live. Missing for 10: independent/hands-on third-party confirmation of the MCP server working in practice and details on transport/auth setup.", "evidenceIds": [ "avalara-docs-6", "avalara-docs-7", "avalara-docs-8", "avalara-docs-13", "avalara-probe-3" ] }, { "productId": "avalara", "storyId": "agentic-nl-commands", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Avalara ships an official MCP server exposing tax calculation, transaction, and nexus tools that an AI agent can invoke via natural-language requests, plus an AI-powered tax-code recommendation feature that accepts free-text product descriptions and a mention of 'conversational queries' for account analysis. However, there's no first-party chat/NL interface for end users and no independent/hands-on evidence of natural-language command execution in practice. Missing for 10: end-user natural-language UI evidence, independent verification of MCP/NL command reliability, broader NL coverage beyond tax-code suggestion and conversational analytics.", "evidenceIds": [ "avalara-docs-6", "avalara-docs-7", "avalara-docs-12", "avalara-probe-3" ] }, { "productId": "avalara", "storyId": "agentic-official-cli", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of an official CLI tool for AvaTax; the ecosystem offers REST SDKs (C#, Java, JS, Python, PHP, Ruby) and an MCP server, but nothing about a command-line interface for AI-native workflows.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "agentic-public-api", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Avalara provides a well-documented public REST v2 API with SDKs in six languages, detailed method-level docs (CreateTransaction, CreateNexus), a free trial for API access, and even an MCP server exposing tax calculation, transaction, and nexus endpoints for AI-driven use. missing for 10: no discoverable OpenAPI/swagger spec at standard endpoints (404s), and no independent/hands-on developer corroboration beyond vendor docs.", "evidenceIds": [ "avalara-docs-3", "avalara-docs-4", "avalara-docs-5", "avalara-docs-6", "avalara-docs-9", "avalara-probe-1", "avalara-probe-2", "avalara-probe-3" ] }, { "productId": "avalara", "storyId": "agentic-scoped-keys", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence of scoped/least-privilege API credentialing (e.g., API keys limited to specific permissions or agent-scoped tokens); documentation covers general API/SDK access, account keys, and MCP server capabilities but nothing about issuing restricted-scope credentials for agents.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "agentic-sdks", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Avalara officially documents SDKs for REST v2 APIs across six languages (C#, Java, JavaScript, Python, PHP, Ruby), with clear API method docs and a free trial for building against them. Missing for 10: independent/hands-on developer corroboration of SDK quality and no public OpenAPI spec discoverable (probe returned 404s).", "evidenceIds": [ "avalara-docs-3", "avalara-docs-9", "avalara-docs-4", "avalara-probe-2" ] }, { "productId": "avalara", "storyId": "agentic-webhooks", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence anywhere in the pack of a webhook/event subscription mechanism for AvaTax; the API/MCP surface described is request-response (transactions, nexus, tax code recommendations) with no push/event notification feature documented.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "ai-taxability-answers", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows AvaTax exposes APIs/MCP tools for tax calculation, nexus lookup, and AI-powered tax-code suggestions, plus a generic 'conversational queries' feature for transaction analytics — but nothing describes a first-party AI assistant that answers 'is this taxable in this state' or 'why did this invoice tax this way' with citations to its rule base. The MCP/API surface is designed for external agents/integrations to consume, not a built-in citation-grounded tax Q&A product.", "evidenceIds": [ "avalara-docs-6", "avalara-docs-7", "avalara-docs-12", "avalara-docs-13" ] }, { "productId": "avalara", "storyId": "api-interactive-docs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence shows only static API method documentation pages (e.g., CreateTransaction, CreateNexus) and SDK lists, with no mention of an interactive 'try it' console or runnable code examples. A probe for OpenAPI/Swagger specs (which typically power interactive references) returned 404s across all candidate paths, and no other evidence points to an interactive explorer.", "evidenceIds": [ "avalara-docs-4", "avalara-docs-5", "avalara-probe-2" ] }, { "productId": "avalara", "storyId": "api-machine-spec", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows explicit probing for common OpenAPI/Swagger spec locations (openapi.json, swagger.json, etc.) all returning 404, and no documentation item claims a downloadable OpenAPI/Swagger file for the REST v2 API; only SDKs and API method docs are mentioned.", "evidenceIds": [ "avalara-probe-2", "avalara-docs-3" ] }, { "productId": "avalara", "storyId": "api-sandbox", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Avalara offers a 90-day free trial with API/SDK access without a credit card, which functions as a sandbox for testing without touching production billing data, and the MCP server exposes transaction/nexus endpoints that could be used against trial accounts. However, there's no explicit documentation of a dedicated 'sandbox' environment/endpoint distinct from trial accounts, no mention of sandbox-specific flags or test-mode transaction handling, and no independent confirmation of how cleanly sandbox/trial data is isolated from production. missing for 10: explicit sandbox environment documentation, test-mode API parameters, independent verification of production/sandbox data isolation.", "evidenceIds": [ "avalara-docs-9", "avalara-docs-6", "avalara-docs-4" ] }, { "productId": "avalara", "storyId": "api-versioning-policy", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence confirms AvaTax exposes versioned REST APIs (e.g., 'REST v2 APIs') but nowhere in the pack is there a documented deprecation policy, versioning changelog, or sunset timeline for API versions. missing for 10: explicit deprecation/versioning policy documentation, changelog of API version lifecycle, sunset notice process.", "evidenceIds": [ "avalara-docs-3", "avalara-docs-4" ] }, { "productId": "avalara", "storyId": "autofiling-remittance", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers tax calculation, nexus determination, transaction recording, and exemption certificate management, but contains no documentation of Avalara's actual returns-filing/remittance service or its on-time filing guarantee. A community comment even notes that per-jurisdiction registration and filing remain separate manual work outside of tax calculation tools like Avalara, reinforcing the absence of evidence for automated filing.", "evidenceIds": [ "avalara-comm-1" ] }, { "productId": "avalara", "storyId": "automation-bulk-operations", "verdict": "partial", "quality": 2, "confidence": "low", "rationale": "The only hint of multi-item processing is the Nexus API supporting creation of 'one or more' declarations in a call; the MCP and REST docs otherwise describe single-transaction creation and per-item tax-code suggestions (1-3 per query), with no documented batch/bulk transaction, bulk calculation, or bulk import endpoint for AI-native workflows. Missing for 10: a dedicated batch/bulk transaction API, bulk calculation across many line items, and evidence of AI agents invoking such bulk operations via MCP.", "evidenceIds": [ "avalara-docs-5", "avalara-docs-4", "avalara-docs-7", "avalara-docs-8" ] }, { "productId": "avalara", "storyId": "automation-rules-engine", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence shows AvaTax as an API/MCP-driven tax calculation, transaction, and nexus system, but there is no evidence of a user-facing rules engine where AI-native users can define custom event-triggered automations beyond calling discrete API/MCP endpoints. missing for 10: evidence of a rule-definition interface, event-trigger configuration, or workflow automation engine tied to conditions and actions.", "evidenceIds": [ "avalara-docs-4", "avalara-docs-5", "avalara-docs-6", "avalara-docs-13" ] }, { "productId": "avalara", "storyId": "automation-scheduled-jobs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers tax calculation, transaction APIs, nexus, SDKs, and an MCP server, but nothing describes scheduling recurring jobs/workflows or any automation/orchestration scheduler capability. Missing for 10: any documentation of recurring job scheduling, workflow automation triggers, or cron-like execution features.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "automation-versioned-workflows", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of version control, review workflows, or rollback capabilities for AvaTax automations/configurations; the evidence covers tax calculation APIs, SDKs, and an MCP server but nothing about versioning or reverting automation logic.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "billing-stack-integrations", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Evidence shows Avalara exposes REST APIs, SDKs in multiple languages, and a CreateTransaction endpoint that a developer could use to build tax integration into a billing or commerce system, but there is no evidence of maintained, pre-built connectors for specific payment processors, billing systems, or commerce platforms (e.g., Shopify, Stripe, NetSuite plugins) that would let transactions flow in without custom plumbing. Missing for 10: named pre-built platform/plugin integrations, connector marketplace, or documentation of maintained third-party connectors for billing/payment/commerce systems.", "evidenceIds": [ "avalara-docs-3", "avalara-docs-4", "avalara-docs-6", "avalara-docs-9" ] }, { "productId": "avalara", "storyId": "calculation-accuracy-guarantee", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence in the pack mentions an accuracy guarantee, indemnification, or audit-defense commitment for calculations produced by AvaTax; documentation focuses on API/SDK capabilities and coverage, not financial/legal backing of results.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "compliant-invoice-fields", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "AvaTax calculates jurisdiction-aware tax, validates addresses, and returns transaction-level tax data (avalara-docs-1, avalara-docs-2, avalara-docs-4) that a system could use to populate invoice fields, and nexus data ties to registration obligations (avalara-docs-5, avalara-docs-13). However there is no evidence of AvaTax itself generating compliant invoice documents with required VAT fields/registration numbers or locale-specific invoice formatting — it's positioned as a calculation/transaction API, and community commentary notes registration/filing remain separate manual work (avalara-comm-1). Missing for 10: evidence of actual invoice document generation or field mapping guidance for VAT-compliant invoices, registration number placement on invoices, and locale-specific invoice templates.", "evidenceIds": [ "avalara-docs-1", "avalara-docs-2", "avalara-docs-4", "avalara-docs-5", "avalara-docs-13", "avalara-comm-1" ] }, { "productId": "avalara", "storyId": "exempt-customer-api", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "AvaTax's CreateTransaction API supports recording transactions and Exemption Certificate Management (ECM/CertCapture) exists to collect/validate/store certificates, suggesting the capability exists, but the evidence never shows an explicit exemption/exemptCertNo parameter on the transaction API nor documents how the exemption reason flows through to reporting and filed returns. missing for 10: explicit exemption-reason field/parameter in the transaction API, documented linkage between ECM certificates and CreateTransaction, and evidence the reason appears in filed-return output.", "evidenceIds": [ "avalara-docs-4", "avalara-docs-11" ] }, { "productId": "avalara", "storyId": "exemption-certificate-management", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Avalara's Exemption Certificate Management (ECM/CertCapture) is documented to collect, validate, and store exemption certificates in one place, and AvaTax's tax calculation API is the natural integration point, but the evidence pack never explicitly documents expiry tracking or automatic application of stored certificates at calculation time. missing for 10: explicit documentation of certificate expiry/renewal tracking, and explicit evidence that AvaTax calculation automatically applies stored certificate exemptions without manual steps.", "evidenceIds": [ "avalara-docs-11", "avalara-docs-4", "avalara-docs-1" ] }, { "productId": "avalara", "storyId": "filing-calendar-visibility", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers AvaTax's calculation, transaction recording, nexus, and MCP tooling, but contains no mention of a filing calendar, per-filing status tracking, remitted amounts, or downloadable filing confirmations. A community comment even notes that per-jurisdiction registration and filing remain separate work outside core AvaTax, reinforcing that this dashboard-style filing visibility isn't evidenced here.", "evidenceIds": [ "avalara-docs-4", "avalara-docs-5", "avalara-comm-1" ] }, { "productId": "avalara", "storyId": "global-vat-gst", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Avalara's docs confirm global coverage (200+ countries/jurisdictions) and jurisdiction-aware tax calculation, nexus tracking, and transaction management APIs, which supports VAT/GST scenarios beyond US sales tax. However, the evidence pack doesn't explicitly confirm EU OSS/IOSS scheme support, UK-specific VAT filing, or automated foreign registration threshold monitoring/alerts, and a community source notes that registration and filing remain separate manual work even when tax calculation is handled. missing for 10: explicit EU OSS/IOSS scheme documentation, UK/Canada/Australia-specific filing support details, automated foreign registration threshold tracking/alerts, and independent confirmation that filings are actually supported end-to-end rather than requiring separate manual registration/filing work.", "evidenceIds": [ "avalara-docs-1", "avalara-docs-5", "avalara-docs-10", "avalara-docs-13", "avalara-comm-1" ] }, { "productId": "avalara", "storyId": "historical-exposure-analysis", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows nexus determination, transaction creation/retrieval, and tax calculation APIs, but nothing about importing historical transactions to produce a backdated exposure report with estimated liability, penalties, and interest per jurisdiction. missing for 10: historical transaction import/backfill tooling, exposure/liability estimation engine, penalty and interest calculation, jurisdiction-level backdated exposure reporting.", "evidenceIds": [ "avalara-docs-5", "avalara-docs-13", "avalara-docs-4", "avalara-docs-8" ] }, { "productId": "avalara", "storyId": "liability-reporting", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "AvaTax exposes transaction and nexus data via API/MCP that could underpin liability tracking, but there is no direct evidence of a jurisdiction-level accrued liability report (collected vs owed), filing-ready summaries, or GL/accountant-ready export formats. Community input even notes that filing/registration remains separate manual work outside the core tax calc engine. missing for 10: dedicated liability/reconciliation report, GL export format documentation, collected-vs-owed dashboard evidence.", "evidenceIds": [ "avalara-docs-8", "avalara-docs-13", "avalara-docs-6", "avalara-comm-1" ] }, { "productId": "avalara", "storyId": "marketplace-facilitator-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers general tax calculation, transaction recording, nexus declarations, and jurisdiction mapping, but nothing addresses marketplace facilitator scenarios specifically — e.g., flagging transactions where Amazon or an app store already collected/remitted tax so they count toward nexus thresholds without being taxed/remitted again. This is a standard, fair axis for a tax compliance product, but no documentation, API method, or feature description in the pack mentions facilitator-collected tax, marketplace facilitator laws, or non-remittance flags.", "evidenceIds": [ "avalara-docs-4", "avalara-docs-5", "avalara-docs-13" ] }, { "productId": "avalara", "storyId": "nexus-threshold-monitoring", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence covers nexus declaration (CreateNexus), retrieving existing nexus definitions, and transaction recording/calculation, but nothing shows AvaTax proactively tracking sales/transaction counts against state/country thresholds or alerting a founder before a threshold is crossed. CreateNexus appears to be for declaring nexus you've already determined, not predictive monitoring. Missing for full/partial: any documented threshold-tracking dashboard, automated alerts, or predictive nexus-crossing notifications.", "evidenceIds": [ "avalara-docs-5", "avalara-docs-13", "avalara-docs-4" ] }, { "productId": "avalara", "storyId": "omnichannel-import", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "AvaTax exposes a transaction API (CreateTransaction, transaction retrieval/filtering via MCP) and multi-language SDKs, which let developers push or backfill transactions programmatically from any custom integration. However, the evidence pack contains no mention of pre-built connectors or native sync integrations with specific storefronts/processors (e.g., Shopify, Stripe, marketplaces) that would let non-developers or bulk imports cover 'every channel' without custom API work. missing for 10: documented connector/integration marketplace, batch/backfill import tooling, evidence of multi-source reconciliation for filings.", "evidenceIds": [ "avalara-docs-4", "avalara-docs-8", "avalara-docs-3", "avalara-docs-6" ] }, { "productId": "avalara", "storyId": "openness-api-parity", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "AvaTax exposes broad API/SDK coverage for core functions (tax calculation, transactions, nexus, exemption certs) and even an MCP server with AI-oriented tools like tax code recommendations, suggesting most UI functionality is API-accessible. However, there's no evidence confirming full parity with the UI (e.g., account administration, reporting/dashboards, user management) and no public OpenAPI spec was found via probe, limiting verifiable completeness. missing for 10: explicit UI-to-API parity confirmation, published OpenAPI/spec for full endpoint coverage, evidence covering admin/reporting UI features, independent verification of parity claims.", "evidenceIds": [ "avalara-docs-3", "avalara-docs-4", "avalara-docs-5", "avalara-docs-6", "avalara-docs-7", "avalara-docs-8", "avalara-docs-11", "avalara-docs-13", "avalara-probe-2", "avalara-probe-3" ] }, { "productId": "avalara", "storyId": "openness-full-export", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "AvaTax exposes APIs to retrieve individual transactions and nexus records, but the evidence pack contains no mention of bulk/full data export, open-format dumps, or any documented data-portability/exit path for exemption certificates, transaction history, or nexus configurations.", "evidenceIds": [ "avalara-docs-8", "avalara-docs-13", "avalara-docs-11" ] }, { "productId": "avalara", "storyId": "openness-open-license", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "AvaTax is a proprietary closed-source commercial SaaS/API product with SDKs and docs, but there is no evidence of the product's source code being available under an open license; SDK availability is not the same as open-sourcing the underlying tax engine/service.", "evidenceIds": [ "avalara-docs-3" ] }, { "productId": "avalara", "storyId": "openness-self-host", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "AvaTax is a cloud-hosted SaaS tax-calculation API/service, not open-source software distributed for self-hosting; self-hosting the 'core product' is a category error for a proprietary compliance/data service that relies on centrally maintained tax rules and jurisdiction data.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "privacy-data-residency", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence in the pack addresses data residency, regional data storage options, or data localization controls for AvaTax; the evidence only covers tax calculation, APIs, SDKs, and MCP tooling. This is a fair axis to ask about a SaaS product handling transaction data, so absence of evidence yields 'none' rather than 'na'.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "privacy-no-training", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "AvaTax is a tax-calculation API/service, not an AI model provider or platform where user data is used for AI training; no AI training data-usage controls or opt-outs are a relevant axis for this product category.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "privacy-retention-controls", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence describes data retention controls, deletion APIs, or privacy/compliance settings for AvaTax data; the pack only covers tax calculation, nexus, transactions, and MCP tool access. Missing for 10: any documentation of data retention policy, deletion/right-to-be-forgotten API, or configurable data lifecycle controls.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "privacy-telemetry-optout", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence in the pack addresses telemetry/usage-tracking opt-out settings for AvaTax; the docs cover tax calculation, SDKs, MCP tools, and pricing but say nothing about data collection or privacy controls for AI-native usage.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "product-tax-codes", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "AvaTax's core tax-code system lets developers assign a taxability classification (SaaS, digital goods, services, physical goods, etc.) to each catalog item, and the docs show an AI-powered tax-code recommendation tool that returns ranked tax code suggestions from product description/category/itemType, plus transaction creation APIs that use these codes for jurisdiction-aware calculation across 200+ countries. Missing for 10: no explicit documentation of the full tax-code taxonomy/category list or a bulk catalog-mapping workflow example, and no independent/hands-on validation of mapping accuracy.", "evidenceIds": [ "avalara-docs-7", "avalara-docs-4", "avalara-docs-1", "avalara-docs-10", "avalara-docs-13" ] }, { "productId": "avalara", "storyId": "rate-rule-currency", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Docs confirm AvaTax is a cloud API that calculates tax using 'current tax rules and boundaries' across 200+ jurisdictions, implying vendor-managed rate/rule updates without client redeploys (avalara-docs-1, avalara-docs-10). However, there's no explicit documentation of an update changelog, notification cadence, or effective-date tracking for jurisdiction rule changes. Missing for 10: documented release notes/changelog for rate changes, explicit statement of update frequency, and confirmation of no-redeploy guarantee beyond implication of being an API service.", "evidenceIds": [ "avalara-docs-1", "avalara-docs-10", "avalara-docs-13" ] }, { "productId": "avalara", "storyId": "realtime-tax-calculation", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "Docs explicitly describe real-time tax calculation during checkout via the CreateTransaction API, with address validation and jurisdiction mapping for rooftop-level accuracy, plus multi-language SDKs for integration into checkout flows. Missing for 10: independent benchmark or hands-on evidence of sub-checkout-latency performance, and explicit confirmation of 'rooftop' (parcel-level) precision versus jurisdiction-level rounding.", "evidenceIds": [ "avalara-docs-1", "avalara-docs-2", "avalara-docs-4", "avalara-docs-3", "avalara-docs-10" ] }, { "productId": "avalara", "storyId": "refund-adjustment-handling", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "The evidence pack covers transaction creation, nexus, address validation, and MCP tools, but nothing describes refund/credit-note transaction types, negative-amount handling, or how such adjustments flow into liability reporting or subsequent filings. Missing for 10: documentation of refund/credit-note transaction codes (e.g., ReturnInvoice), negative-amount liability adjustment logic, and evidence that filings/reports automatically reconcile these without manual journal entries.", "evidenceIds": [] }, { "productId": "avalara", "storyId": "saas-digital-taxability", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "AvaTax provides jurisdiction-aware tax calculation, tax code recommendations, and nexus data via API/MCP, which supports differentiated SaaS/digital-goods taxability determinations, but there is no evidence of visible research/rationale behind each determination, no documented handling of bundled-offering taxability logic, and no SaaS-specific taxability guidance. missing for 10: transparent research/citations behind each taxability determination, explicit bundled-offering taxability rules, SaaS-specific taxability examples, independent validation of accuracy on hard cases.", "evidenceIds": [ "avalara-docs-1", "avalara-docs-7", "avalara-docs-13", "avalara-docs-10" ] }, { "productId": "avalara", "storyId": "state-registrations", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack only covers AvaTax's tax calculation, transaction recording, nexus tracking, and MCP/API capabilities — none of it describes an in-product service where Avalara files state/local sales tax permit registrations on the customer's behalf with per-registration pricing or status tracking. The nexus API merely tracks declared nexus, it does not file registrations. Community evidence (avalara-comm-1) explicitly notes that per-jurisdiction tax registration and filing remains separate work even when using tools like Avalara, reinforcing that this capability is not delivered in-product.", "evidenceIds": [ "avalara-docs-5", "avalara-docs-13", "avalara-comm-1" ] }, { "productId": "avalara", "storyId": "transaction-audit-trail", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "AvaTax's CreateTransaction and GetTransactions APIs record and retrieve individual transaction details (address, tax amount, presumably jurisdiction) and nexus/tax-code data can be queried, suggesting some reconstructability, but there is no explicit documentation of a consolidated audit trail showing the exact rate, rule, and product-category reasoning per transaction, nor an export mechanism for audit purposes. Missing for 10: explicit audit-log/export feature documentation, evidence of rule/rate versioning retained per transaction, and independent confirmation of audit usability.", "evidenceIds": [ "avalara-docs-4", "avalara-docs-8", "avalara-docs-13", "avalara-docs-2" ] }, { "productId": "avalara", "storyId": "vat-id-validation", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows general tax calculation, jurisdiction mapping, nexus handling, and 200+ country coverage, but nothing documents real-time validation of VAT/tax IDs against official registries (e.g., VIES) or automatic application of reverse-charge logic for B2B cross-border transactions. Missing for 10: VAT ID/registry validation API, reverse-charge determination logic, B2B cross-border specific documentation or examples.", "evidenceIds": [ "avalara-docs-1", "avalara-docs-2", "avalara-docs-5", "avalara-docs-10", "avalara-docs-13" ] }, { "productId": "avalara", "storyId": "vda-remediation", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers nexus tracking, tax calculation, exemption certificates, and transaction APIs, but contains no mention of voluntary disclosure agreements, back-filing, or remediation services for past-due jurisdictions. This is a plausible axis for a tax compliance vendor, but nothing in the pack substantiates it.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "agent-reads-tax-state", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "The REST API and SDKs expose real tax-posture endpoints an agent could call — nexus list, registrations, filings, exemptions, tax estimates — and API keys support expiration-based scoping, so the API half of the story is solidly evidenced. However, Kintsugi's actual MCP server is documented as giving AI assistants access to API documentation for building integrations, not as an interface for reading or acting on live nexus/registration/filing data — so the 'MCP surface' part of the story is not delivered as described, only a docs-lookup MCP exists. Missing for 10: an MCP server (or API scope) that lets an agent directly read/act on nexus, registrations, accrued liability and filings rather than just fetch documentation; fine-grained credential scopes beyond key expiration; independent/hands-on confirmation of agentic use.", "evidenceIds": [ "kintsugi-docs-6", "kintsugi-docs-11", "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-5", "kintsugi-docs-2", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-agent-docs", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "A probe confirms llms.txt is live and served (HTTP 200) with a clear description of the docs, and Kintsugi also documents an official MCP server that gives AI assistants direct access to its API documentation. Together these show agent-oriented docs access is explicitly supported. missing for 10: no independent third-party confirmation of agent usage/quality beyond the vendor's own probe and docs, and no evidence of a machine-readable openapi.json to complement llms.txt (404s on candidate paths).", "evidenceIds": [ "kintsugi-probe-1", "kintsugi-docs-6", "kintsugi-probe-3", "kintsugi-probe-2" ] }, { "productId": "kintsugi", "storyId": "agentic-ai-insights", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Kintsugi's evidence shows nexus/threshold determination logic and an MCP server aimed at helping developers build integrations with accurate API docs, not AI-generated insights or suggestions surfaced to end users from their own data. No feature is described that analyzes user data and proactively generates recommendations or insights within the product UI.", "evidenceIds": [ "kintsugi-docs-3", "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-autonomous-automation", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "Kintsugi's connectors auto-sync transactions and its tax engine autonomously computes nexus, tracks exposure, and can register/file across jurisdictions once configured (kintsugi-docs-1, kintsugi-docs-15), which is background automation, but this is standard SaaS integration behavior rather than a user-defined 'automation' or agentic workflow builder, and the MCP server is scoped only to serving API docs to coding assistants, not orchestrating autonomous tasks (kintsugi-docs-6, kintsugi-probe-3). Missing for 10: evidence of a configurable automation/workflow engine, scheduling or trigger-based agent tasks, and any AI-agent-driven autonomous execution beyond passive data sync.", "evidenceIds": [ "kintsugi-docs-1", "kintsugi-docs-15", "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-builtin-assistant", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Kintsugi's evidence only describes an MCP server that lets external AI assistants read its API documentation for developers building integrations (kintsugi-docs-6, kintsugi-probe-3) — this is not a built-in assistant inside the product that end users can delegate tasks to. No feature resembling an in-app AI assistant, chat, or task-delegation UI is documented anywhere in the pack.", "evidenceIds": [ "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-headless", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Kintsugi is API-first: it exposes a REST API and five official SDKs plus configurable API keys, making it straightforward to call from scripts or CI pipelines without a UI. However, there is no explicit CLI, headless-mode documentation, or CI-specific guidance (e.g., GitHub Actions examples) confirming automated pipeline use. Missing for 10: dedicated CLI/headless tooling, CI/CD integration examples, and explicit 'automation' framing beyond generic API/SDK access.", "evidenceIds": [ "kintsugi-docs-2", "kintsugi-docs-5", "kintsugi-docs-10", "kintsugi-probe-1" ] }, { "productId": "kintsugi", "storyId": "agentic-mcp-client", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a tax-compliance SaaS/API platform, not an AI agent; the only MCP-related evidence shows Kintsugi providing its own MCP server for AI assistants to access its API docs (the server role), not Kintsugi acting as a client that plugs in external MCP servers to use their tools. This client-consumption axis is a category mismatch for a compliance API product.", "evidenceIds": [ "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-mcp-server", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Kintsugi documents an official MCP server that gives AI assistants direct access to its API documentation for building integrations, confirmed by both docs and a probe hit. However, the described purpose is limited to documentation/code-example access rather than full agentic action (e.g., invoking tax calculations, registrations, filings via MCP), and there's no independent/hands-on corroboration of its use. Missing for 10: evidence of MCP-driven actions beyond docs lookup, independent verification of setup/usage, and details on tool/action coverage via the server.", "evidenceIds": [ "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-nl-commands", "verdict": "partial", "quality": 3, "confidence": "medium", "rationale": "Kintsugi ships an official MCP server, but per its own docs it only gives AI assistants access to API documentation/code examples for building integrations—it does not let a user perform actual operations (sync transactions, create registrations, file exemptions) via natural-language commands. Missing for 10: an MCP/agent interface that executes real product actions (nexus checks, filings, registrations) via natural language, evidence of end-to-end task completion through NL commands, independent corroboration of agentic use.", "evidenceIds": [ "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-official-cli", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "Kintsugi offers SDKs, REST API, and an MCP server for docs access, but no evidence of an official CLI tool being published or documented anywhere in the evidence pack.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "agentic-public-api", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Kintsugi documents a REST API with numerous endpoints (nexus, registrations, filings, exemptions, tax estimate), five official SDKs generated from an OpenAPI spec, API key management, and an llms.txt for AI-native discovery, plus an MCP server exposing docs to AI assistants. Missing for 10: a publicly reachable OpenAPI/swagger spec (probe found all candidate paths 404) and independent third-party corroboration of API robustness.", "evidenceIds": [ "kintsugi-docs-2", "kintsugi-docs-9", "kintsugi-docs-10", "kintsugi-docs-11", "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-14", "kintsugi-docs-5", "kintsugi-docs-6", "kintsugi-probe-1", "kintsugi-probe-2", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "agentic-scoped-keys", "verdict": "partial", "quality": 4, "confidence": "medium", "rationale": "Kintsugi lets users create API keys with configurable expiration windows (Never/1mo/6mo/1yr), which supports time-boxing credentials, but there's no evidence of scoped/permission-based least-privilege controls (e.g., read-only vs write, endpoint-level restrictions) for issuing to an agent specifically. The MCP server only exposes API docs to assistants, not credential-scoping. Missing for 10: permission/role-based scoping of API keys, agent-specific credential issuance workflow, and documentation tying key creation to least-privilege agent use cases.", "evidenceIds": [ "kintsugi-docs-5", "kintsugi-docs-6" ] }, { "productId": "kintsugi", "storyId": "agentic-sdks", "verdict": "full", "quality": 8, "confidence": "medium", "rationale": "Docs explicitly confirm five official SDKs (Python, TypeScript, Java, PHP, Ruby) generated from a shared OpenAPI spec, plus REST API docs and an official MCP server for AI-assisted integration. Missing for 10: independent/hands-on corroboration of SDK quality and a publicly reachable OpenAPI spec (probe found 404s on candidate URLs).", "evidenceIds": [ "kintsugi-docs-10", "kintsugi-docs-2", "kintsugi-docs-6", "kintsugi-probe-3", "kintsugi-probe-2" ] }, { "productId": "kintsugi", "storyId": "agentic-webhooks", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence pack item mentions webhooks, event subscriptions, or push notifications; Kintsugi's documented integration model is REST API/SDK polling and connector sync, not event-driven webhooks.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "ai-taxability-answers", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows Kintsugi's MCP server exposes API documentation for developers building integrations, not an AI agent answering founder-facing tax questions like nexus or invoice-taxation rationale with citations to a rule base. No docs describe a conversational/Q&A tax-explanation feature.", "evidenceIds": [ "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "api-interactive-docs", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Kintsugi has structured API reference pages (get-nexus-for-org, create-registration, get-filings, create-exemption) and SDKs generated from an OpenAPI spec, suggesting a formal reference exists, but no evidence explicitly confirms an interactive 'try it out' console or runnable code examples, and probes for a public openapi.json spec returned 404s. missing for 10: explicit confirmation of in-browser runnable/try-it examples, working public OpenAPI spec endpoint, independent corroboration of interactivity.", "evidenceIds": [ "kintsugi-docs-10", "kintsugi-docs-11", "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-14", "kintsugi-probe-2" ] }, { "productId": "kintsugi", "storyId": "api-machine-spec", "verdict": "disputed", "quality": 3, "confidence": "medium", "rationale": "Docs assert an OpenAPI spec underlies the SDKs and API reference (kintsugi-docs-10), but a direct probe for standard OpenAPI/Swagger download endpoints (openapi.json, swagger.json, etc.) returned 404 everywhere (kintsugi-probe-2), so no machine-readable spec is actually downloadable. Missing for 10: a working, publicly downloadable OpenAPI/Swagger file or documented export endpoint, and independent confirmation of successful spec retrieval.", "evidenceIds": [ "kintsugi-docs-10", "kintsugi-probe-2" ] }, { "productId": "kintsugi", "storyId": "api-sandbox", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence pack mention of a sandbox/test environment, test API keys, or staging mode distinct from production; docs cover connectors, SDKs, nexus, and MCP for docs access but nothing about isolated testing without touching production data.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "api-versioning-policy", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows a REST API, SDKs generated from an OpenAPI spec, and API key management, but no mention anywhere of API versioning scheme or a documented deprecation policy; the OpenAPI spec itself couldn't even be located via probing. missing for 10: explicit API version numbering/headers, a deprecation/sunset policy document, changelog or migration guidance for breaking changes.", "evidenceIds": [ "kintsugi-docs-2", "kintsugi-docs-10", "kintsugi-probe-2" ] }, { "productId": "kintsugi", "storyId": "autofiling-remittance", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Docs confirm registration creation, filings retrieval, and marketing copy claiming Kintsugi 'registers you and files everywhere you sell,' indicating automated filing/remittance across jurisdictions. However, there is no explicit documentation of an on-time filing guarantee or penalty-coverage commitment, nor independent evidence corroborating the claim. missing for 10: explicit vendor guarantee/penalty-coverage terms, evidence of remittance execution (not just filing), independent/hands-on confirmation of on-time filing performance.", "evidenceIds": [ "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-15" ] }, { "productId": "kintsugi", "storyId": "automation-bulk-operations", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Kintsugi's API supports syncing large volumes of transaction data (a full year's worth) and offers paginated list endpoints (e.g., Get Filings, Get Nexus for Org), which implies handling many records, but there is no documented bulk/batch endpoint for creating or updating multiple exemptions, registrations, or other items in a single call. Missing for 10: explicit bulk-create/bulk-update API endpoints, batch operation examples, and evidence of an AI agent orchestrating multi-item actions via MCP or SDK.", "evidenceIds": [ "kintsugi-docs-3", "kintsugi-docs-11", "kintsugi-docs-13", "kintsugi-docs-2" ] }, { "productId": "kintsugi", "storyId": "automation-rules-engine", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a sales-tax compliance API with connectors, SDKs, and an MCP server for documentation access, but there is no evidence of a rules engine or event-triggered automation system where users define conditional actions on events. The MCP server only exposes API docs, not workflow/rule automation.", "evidenceIds": [ "kintsugi-docs-6", "kintsugi-docs-2" ] }, { "productId": "kintsugi", "storyId": "automation-scheduled-jobs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Kintsugi's docs describe automatic background transaction syncing and filing, but there is no evidence of an API, SDK, or MCP capability that lets an AI-native user explicitly schedule or configure recurring jobs/workflows (e.g., cron-like triggers, webhook-based automation, or workflow orchestration). The MCP server only exposes documentation access, not job scheduling. Missing for 10: any documented scheduling/automation API, recurring job configuration options, or workflow-orchestration tooling for AI agents.", "evidenceIds": [ "kintsugi-docs-1", "kintsugi-docs-6", "kintsugi-probe-3" ] }, { "productId": "kintsugi", "storyId": "automation-versioned-workflows", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a sales-tax compliance API/platform, not an automation-building tool with workflows to version, review, or roll back; this axis is a category error for this product type.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "billing-stack-integrations", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Kintsugi advertises dozens of prebuilt connectors for shopping carts, billing systems, ERPs, and accounting platforms, plus REST API/SDKs for custom integrations, which supports the 'no plumbing' claim. However, the evidence pack lacks a named list of specific maintained integrations (e.g., Stripe, Shopify) or independent confirmation of connector reliability/maintenance status. Missing for 10: a documented list of specific named connectors with maintenance status, and independent/hands-on validation that these prebuilt integrations work as advertised.", "evidenceIds": [ "kintsugi-docs-1", "kintsugi-docs-2", "kintsugi-docs-10" ] }, { "productId": "kintsugi", "storyId": "calculation-accuracy-guarantee", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of an accuracy guarantee, audit support commitment, or liability coverage for incorrect tax calculations anywhere in the docs; the pack covers integrations, API endpoints, and connectors but nothing about accuracy warranties or audit defense.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "compliant-invoice-fields", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence describes Kintsugi as a sales-tax/nexus compliance API (calculating tax, tracking nexus, registrations, filings, exemptions) but never mentions VAT-specific invoice requirements, required VAT fields, registration-number formatting, or locale invoicing rules for non-US jurisdictions.", "evidenceIds": [ "kintsugi-docs-9", "kintsugi-docs-11", "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-14" ] }, { "productId": "kintsugi", "storyId": "exempt-customer-api", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Kintsugi documents a Create Exemption API that lets developers define exemption type, jurisdiction, and validity dates via the API (kintsugi-docs-14), and the tax estimate/calculation endpoint exists separately (kintsugi-docs-9), implying exemptions feed into tax calc. However there is no explicit documentation showing the exemption reason is carried through calculation, reporting dashboards, and the actual filed return — the pack never ties exemption records to reporting or filing outputs. Missing for 10: evidence that exemption reason propagates into tax reports and filed returns, and confirmation it applies per-transaction or per-customer at calculation time.", "evidenceIds": [ "kintsugi-docs-14", "kintsugi-docs-9", "kintsugi-docs-13" ] }, { "productId": "kintsugi", "storyId": "exemption-certificate-management", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Kintsugi has a Create Exemption API allowing exemption records with type, jurisdiction, and validity dates to be stored (kintsugi-docs-14), but there is no evidence of a customer-facing collection workflow (e.g., certificate upload portal), automated validation of certificate authenticity, expiry-based alerting, or documented automatic application of exemptions at tax-calculation time. missing for 10: customer-facing certificate collection/upload flow, validation logic against jurisdiction rules, expiry tracking/alerts, and documented automatic application of stored exemptions during tax calculation.", "evidenceIds": [ "kintsugi-docs-14", "kintsugi-docs-9" ] }, { "productId": "kintsugi", "storyId": "filing-calendar-visibility", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Kintsugi's Get Filings API supports listing filings with status/date/jurisdiction filters, and Create Registration tracks filings across jurisdictions, giving some per-filing status visibility, but there is no evidence of a calendar view of upcoming returns, remitted amounts, or downloadable filing confirmations. missing for 10: calendar UI for upcoming returns, remittance amount tracking, downloadable confirmation documents for filed returns.", "evidenceIds": [ "kintsugi-docs-13", "kintsugi-docs-12", "kintsugi-docs-11" ] }, { "productId": "kintsugi", "storyId": "global-vat-gst", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack only describes US-style sales tax nexus, registration, and filing APIs (with generic \"Country\" fields) but never mentions VAT, GST, EU OSS/IOSS, or specific UK/Canada/Australia regimes, so there is no proof Kintsugi extends its compliance engine to international VAT/GST obligations.", "evidenceIds": [ "kintsugi-docs-3", "kintsugi-docs-11", "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-14" ] }, { "productId": "kintsugi", "storyId": "historical-exposure-analysis", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Kintsugi documents historical transaction import spanning the prior calendar year plus current, explicitly for nexus/threshold backdating (kintsugi-docs-3), and markets 'finds the exposure you didn't know you had' with nexus tracking APIs (kintsugi-docs-11, kintsugi-docs-15). However, no evidence describes computing estimated liability, penalties, or interest amounts per jurisdiction — only nexus/threshold detection and registration/filing tracking are documented. Missing for 10: explicit liability/penalty/interest calculation methodology, per-jurisdiction dollar exposure reporting, and any UI/report showing a backdated liability summary.", "evidenceIds": [ "kintsugi-docs-3", "kintsugi-docs-11", "kintsugi-docs-15", "kintsugi-docs-4" ] }, { "productId": "kintsugi", "storyId": "liability-reporting", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Kintsugi provides nexus tracking, filings, registrations, and exemption APIs (get-filings, create-registration, get-nexus-for-org) that could surface liability by jurisdiction, and product categorization determines taxability, but there's no explicit documentation of a 'collected vs owed' liability dashboard or accountant/GL-ready export format. missing for 10: explicit accrued-liability-by-jurisdiction view/report, collected-vs-owed reconciliation feature, documented GL/accountant export formats (e.g. CSV/QBO/NetSuite journal entries), independent confirmation of filing-ready detail accuracy.", "evidenceIds": [ "kintsugi-docs-11", "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-14", "kintsugi-docs-8", "kintsugi-docs-3" ] }, { "productId": "kintsugi", "storyId": "marketplace-facilitator-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence describes handling of marketplace-facilitator-collected transactions (e.g., Amazon, app stores) for nexus tracking without double-remittance; docs cover general nexus threshold tracking, transaction sync, and filings but never mention facilitator tax collection status or how it's flagged/excluded from remittance while still counted toward nexus. Missing for 10: any mention of marketplace facilitator flag/field, guidance on excluding facilitator-collected sales from filings, or documentation of nexus counting facilitator sales without double remittance.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "nexus-threshold-monitoring", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Kintsugi syncs transactions and computes nexus by looking back across the trailing year, and exposes a nexus list endpoint, supporting economic nexus tracking against thresholds — but this evidence is entirely first-party docs with no independent/hands-on corroboration, and none of the pack explicitly describes proactive pre-crossing alerts (only that it 'determines nexus' and 'finds exposure you didn't know you had,' which reads more retrospective than predictive). missing for 10: explicit documentation of real-time threshold alerts/notifications before a crossing occurs, per-state/country threshold configuration detail, and independent confirmation the monitoring works as claimed.", "evidenceIds": [ "kintsugi-docs-3", "kintsugi-docs-11", "kintsugi-docs-15", "kintsugi-docs-1" ] }, { "productId": "kintsugi", "storyId": "omnichannel-import", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Docs confirm prebuilt connectors for carts, billing, ERPs, and accounting platforms plus a REST API/SDKs for backfilling historical transactions (previous full calendar year through today), which supports multi-channel coverage. However, there's no independent/hands-on corroboration of connector breadth or reliability, and OpenAPI spec probes 404, leaving API completeness unverified. Missing for 10: independent verification of connector list/reliability, working public OpenAPI spec, evidence of successful multi-source reconciliation in practice.", "evidenceIds": [ "kintsugi-docs-1", "kintsugi-docs-2", "kintsugi-docs-3", "kintsugi-probe-2" ] }, { "productId": "kintsugi", "storyId": "openness-api-parity", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Kintsugi's REST API/SDKs expose most core tax-compliance actions—nexus lookup, registrations, filings, exemptions, transaction sync, tax estimation, product categorization—that mirror the platform's main UI workflows, and it ships 5 official SDKs plus an MCP server for developer docs access. However there is no explicit documentation claiming full UI/API parity, no accessible OpenAPI spec (probe returned 404s), and management-type UI features (account/user settings, dashboard reporting, filing status changes) are not evidenced via API. Missing for 10: explicit parity statement, public OpenAPI spec, evidence covering all UI-only settings/admin functions.", "evidenceIds": [ "kintsugi-docs-2", "kintsugi-docs-9", "kintsugi-docs-10", "kintsugi-docs-11", "kintsugi-docs-12", "kintsugi-docs-13", "kintsugi-docs-14", "kintsugi-probe-2" ] }, { "productId": "kintsugi", "storyId": "openness-full-export", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence of any data export feature, open-format export, or account portability tooling; the only migration-related content describes importing from Avalara/TaxJar into Kintsugi, not exporting out. This axis is plausible for a SaaS compliance platform, but no supporting evidence exists.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "openness-open-license", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a closed-source SaaS tax-compliance API/platform, not an open-source project; source code availability under an open license is not a fair axis for this kind of product.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "openness-self-host", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a SaaS sales-tax compliance API/platform, not a product designed for local deployment or self-hosting; the evidence pack shows only cloud API, SDKs, and MCP docs access, with no self-hosting concept applicable to this category of hosted compliance service.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "privacy-data-residency", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a sales-tax compliance/API product, not a data-storage platform; data residency/region selection is not a relevant axis for this product category, and no evidence discusses storage location choices.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "privacy-no-training", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a sales-tax compliance API/product; it has no consumer AI-model training relationship with user data, so an 'AI training opt-out' privacy control is a category error for this product type.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "privacy-retention-controls", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a sales-tax compliance API; the evidence pack covers connectors, SDKs, nexus tracking, API keys, and MCP docs, but contains no mention of data retention policies, deletion controls, or privacy/data lifecycle settings for AI-native users.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "privacy-telemetry-optout", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Kintsugi is a sales-tax compliance API/SaaS product, not an AI agent or developer tool with telemetry collection relevant to AI-native usage; no evidence pack content relates to telemetry/usage-tracking opt-out, and this axis is a category error for a tax-calculation platform.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "product-tax-codes", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Kintsugi documents a product categorization system (category + subcategory) that determines tax treatment across jurisdictions, which directly addresses mapping catalog items to tax categories, and this feeds into its estimate/calculation endpoints. However, the evidence pack lacks detail on the actual taxonomy (whether SaaS, digital goods, services, physical goods are explicitly distinct built-in categories), how granular the mapping API is, or independent confirmation of jurisdictional accuracy. Missing for 10: explicit list/taxonomy of category types (SaaS/digital/services/physical), API reference for assigning categories per line item, and independent/hands-on validation of correct jurisdictional taxation.", "evidenceIds": [ "kintsugi-docs-8", "kintsugi-docs-9", "kintsugi-docs-2" ] }, { "productId": "kintsugi", "storyId": "rate-rule-currency", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Kintsugi's architecture is API-based (tax estimate endpoint, product categorization engine) which implies tax logic lives server-side and calls always reflect the vendor's current rules without requiring integrator redeploys, but there is no explicit documentation about how often rates/rules are updated or how such changes are communicated/dated to customers. Missing for 10: a documented update cadence or changelog for jurisdiction rate/rule changes, and any statement that changes take effect with a timestamp/notice to ops users.", "evidenceIds": [ "kintsugi-docs-9", "kintsugi-docs-8", "kintsugi-docs-15" ] }, { "productId": "kintsugi", "storyId": "realtime-tax-calculation", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Docs confirm a REST API (tax estimate endpoint POST /v1/tax/estimate) plus SDKs for real-time tax pricing before payment, and product categorization drives jurisdiction-specific taxation — supporting the core 'calculate tax via API' story. However, there is no explicit evidence of address-level (rooftop) precision versus ZIP-level approximation, and no documented latency/performance benchmarks confirming it's fast enough for checkout. missing for 10: explicit rooftop/address-level accuracy claims, latency/performance guarantees for checkout-speed use, independent verification of calculation accuracy.", "evidenceIds": [ "kintsugi-docs-9", "kintsugi-docs-8", "kintsugi-docs-10", "kintsugi-probe-1" ] }, { "productId": "kintsugi", "storyId": "refund-adjustment-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers transaction syncing, nexus tracking, filings, exemptions, and product categorization, but nothing addresses how refunds or credit notes with negative amounts are processed through liability, reporting, or filings. This is a fair question for a tax compliance API, so the axis applies, but no supporting evidence exists.", "evidenceIds": [] }, { "productId": "kintsugi", "storyId": "saas-digital-taxability", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Kintsugi documents product-category/subcategory classification that determines tax treatment across jurisdictions, implying it handles SaaS/digital-goods taxability differences, but there is no evidence of bundled-offering handling or of visible research/citations backing each determination. missing for 10: bundled-offering tax logic, state-by-state SaaS taxability rules detail, visible research/sourcing behind each determination.", "evidenceIds": [ "kintsugi-docs-8" ] }, { "productId": "kintsugi", "storyId": "state-registrations", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "Kintsugi's marketing copy explicitly claims it 'registers you' for sales tax and files everywhere you sell, and the API exposes a Create Registration endpoint for tracking/managing registrations across jurisdictions, showing in-product registration handling exists. However there is no evidence of transparent per-registration pricing, and status tracking is only inferred from a generic API endpoint rather than documented UI/status flow. Missing for 10: per-registration pricing disclosure, detailed status-tracking workflow/UI evidence, and independent confirmation that Kintsugi actually files the registrations rather than just tracking them.", "evidenceIds": [ "kintsugi-docs-15", "kintsugi-docs-12", "kintsugi-docs-11" ] }, { "productId": "kintsugi", "storyId": "transaction-audit-trail", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Kintsugi documents product-category classification and a tax estimate/calculation endpoint, plus nexus, registration, exemption and filings APIs, but there is no evidence of a retained, per-transaction breakdown of rate/rules/address resolution or an export mechanism for audit reconstruction. missing for 10: documented storage of calculation inputs (rate, rule, address resolution) per transaction, an audit-log or export feature, and any mention of reconstructability for compliance audits.", "evidenceIds": [ "kintsugi-docs-8", "kintsugi-docs-9", "kintsugi-docs-11", "kintsugi-docs-13" ] }, { "productId": "kintsugi", "storyId": "vat-id-validation", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows Kintsugi handling US-style sales tax, nexus determination, exemptions, and filings, but contains no mention of VAT ID validation against official registries (e.g., VIES), reverse-charge mechanism handling, or any EU/cross-border B2B VAT-specific logic. This is an applicable axis for a tax-compliance product, but no evidence demonstrates it.", "evidenceIds": [ "kintsugi-docs-3", "kintsugi-docs-4", "kintsugi-docs-9", "kintsugi-docs-11", "kintsugi-docs-14" ] }, { "productId": "kintsugi", "storyId": "vda-remediation", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Kintsugi's marketing claims it 'finds the exposure you didn't know you had, registers you, and files everywhere you sell,' and nexus lookback plus registration APIs support identifying past-due exposure and creating registrations, but there is no explicit mention of voluntary disclosure agreements (VDAs) or back-filing of historical periods to clean up past liability. missing for 10: explicit VDA program or negotiation support, back-filing of historical/past-due returns, penalty/interest abatement guidance, any case study or documentation describing remediation of already-accrued exposure rather than just prospective registration and filing.", "evidenceIds": [ "kintsugi-docs-15", "kintsugi-docs-3", "kintsugi-docs-12", "kintsugi-docs-11" ] }, { "productId": "numeral", "storyId": "agent-reads-tax-state", "verdict": "full", "quality": 8, "confidence": "medium", "rationale": "Numeral documents a dedicated MCP server exposing nexus status, filings, registrations, transactions, and exemption certificates as agent-callable tools, plus a full REST API for programmatic access to tax posture and actions (calculations, transactions, certificates). This directly matches the story of an agent reading and acting on tax posture via API/MCP. missing for 10: no explicit documentation of scoped/least-privilege credential support for the MCP endpoint or API keys, and no independent/hands-on corroboration of the MCP server's real-world reliability.", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-docs-5", "numeral-docs-3", "numeral-probe-3" ] }, { "productId": "numeral", "storyId": "agentic-agent-docs", "verdict": "full", "quality": 9, "confidence": "high", "rationale": "A live probe confirms llms.txt returns HTTP 200 with a structured doc index at docs.numeral.com/llms.txt, and Numeral also documents agent-oriented resources like Claude skills and an MCP server for AI agents to consult its docs/API. Missing for 10: independent third-party confirmation of an agent successfully using llms.txt in practice.", "evidenceIds": [ "numeral-probe-1", "numeral-docs-6", "numeral-docs-7" ] }, { "productId": "numeral", "storyId": "agentic-ai-insights", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Numeral's MCP server exposes nexus, filings, and transaction data as callable tools, enabling an external AI agent to answer questions like 'which states have I triggered nexus in' — a form of AI-derived insight from the data. However, this requires a separate AI agent/client to consume the MCP tools rather than a built-in, native in-product insights/suggestions feature. Missing for 10: evidence of insights or suggestions generated natively within Numeral's own UI/dashboard without requiring an external agent, and any proactive suggestion capability (vs. query-response only).", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10" ] }, { "productId": "numeral", "storyId": "agentic-autonomous-automation", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Numeral's core product is autonomous background compliance automation (nexus monitoring, filing, remittance, mail handling, certificate tracking) rather than a user-configurable automation builder, and the MCP server lets an AI agent query/act on nexus, filings, and transactions autonomously. However there's no explicit evidence of scheduled/triggered 'set it and forget it' automations that a user configures themselves via the MCP or API (e.g., recurring jobs, webhooks, or event-driven triggers) — missing for 10: documented scheduling/trigger mechanism for autonomous runs, user-configurable automation workflows, and independent confirmation the MCP tools can run unattended in production.", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-docs-11", "numeral-docs-12", "numeral-docs-13" ] }, { "productId": "numeral", "storyId": "agentic-builtin-assistant", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Numeral exposes an MCP server so external AI agents (e.g., Claude) can query its tax data, and offers Claude skills for coding agents integrating its API — but this is exposing Numeral as a tool for outside agents, not a built-in AI assistant living inside Numeral's own product UI that a user delegates tasks to. No evidence of an in-product chat/assistant experience is provided.", "evidenceIds": [ "numeral-docs-6", "numeral-docs-7", "numeral-docs-9" ] }, { "productId": "numeral", "storyId": "agentic-headless", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Numeral is fundamentally a REST API (create products, calculations, transactions via POST calls) which is inherently callable headlessly from any automation pipeline including CI, and there's also an MCP server exposing account operations as callable tools. However, there is no explicit documentation of CI integration, headless CLI, or automated testing/deployment workflows. missing for 10: explicit CI/CD examples or guidance, a CLI tool for scripted/headless runs, and confirmation of non-interactive auth flows suitable for automation.", "evidenceIds": [ "numeral-docs-2", "numeral-docs-3", "numeral-docs-7", "numeral-probe-1" ] }, { "productId": "numeral", "storyId": "agentic-mcp-client", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Numeral is a tax-compliance SaaS/API product that exposes its own data as an MCP server for external AI agents to call (docs-7, docs-9, probe-3) — it is not itself an agent or agent-runtime that would consume other MCP servers' tools. The 'plug MCP servers into this product' framing is a category error for a backend compliance platform; no evidence pack item describes Numeral acting as an MCP client.", "evidenceIds": [ "numeral-docs-7", "numeral-probe-3" ] }, { "productId": "numeral", "storyId": "agentic-mcp-server", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Numeral documents an official MCP server exposing nexus, filings, registrations, transactions, sales, products, and exemption certificates as tools an AI agent can call directly, with specific tool documentation and example use cases. Missing for 10: independent/hands-on corroboration beyond first-party docs, and details on authentication/rate limits for the MCP endpoint.", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-probe-3" ] }, { "productId": "numeral", "storyId": "agentic-nl-commands", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Numeral ships a documented MCP server exposing nexus, filings, transactions, and exemption certificates as tools an AI agent can call, explicitly enabling natural-language queries like 'which states have I triggered nexus in' — this is the core enabler for NL-driven operation. However, this is agent-mediated (via MCP tool calls) rather than a native NL command interface within Numeral's own product UI, and there's no evidence of a built-in chat/command box or independent hands-on validation of the NL experience. Missing for 10: first-party NL interface within the product itself, broader coverage of all actions via NL, independent corroboration of real-world usage.", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-probe-3" ] }, { "productId": "numeral", "storyId": "agentic-official-cli", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence covers a REST API, MCP server, and Claude skills, but no official CLI tool is mentioned anywhere in the docs or probes.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "agentic-public-api", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Numeral publishes a versioned REST API with documented endpoints (products, transactions, calculations, exemption certificates) and a quickstart guide, plus an llms.txt confirming machine-readable API docs are live (HTTP 200). This is a genuine documented public API an AI-native user could drive directly. Missing for 10: a discoverable OpenAPI/Swagger spec (all candidate paths 404), independent third-party corroboration of API usage.", "evidenceIds": [ "numeral-docs-2", "numeral-docs-3", "numeral-docs-4", "numeral-docs-5", "numeral-probe-1", "numeral-probe-2" ] }, { "productId": "numeral", "storyId": "agentic-scoped-keys", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence Numeral offers scoped or least-privilege API key/credential issuance for agents; the docs mention a general MCP server exposing the whole account (nexus, filings, transactions, etc.) but nothing about restricting agent permissions or issuing limited-scope tokens. Missing for 10: any mention of API key scopes, role-based permissions, or agent-specific credential issuance.", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10" ] }, { "productId": "numeral", "storyId": "agentic-sdks", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Numeral offers a documented REST API with detailed object/endpoint docs, an MCP server, and Claude skills for AI agent integration, but there's no evidence of official language-specific SDKs (e.g., Python/JS/Ruby client libraries) — only raw API/docs and an OpenAPI probe that 404s. missing for 10: official multi-language SDK packages, published OpenAPI spec confirming SDK generation, independent developer corroboration of SDK usage.", "evidenceIds": [ "numeral-docs-2", "numeral-docs-6", "numeral-docs-7", "numeral-probe-1", "numeral-probe-2" ] }, { "productId": "numeral", "storyId": "agentic-webhooks", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "Evidence covers MCP server, REST API objects, and integrations, but there is no mention of webhooks or event subscriptions anywhere in the docs pack. This is a plausible axis for an API-driven tax platform, but absence of evidence means it must be marked none.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "ai-taxability-answers", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Numeral ships an MCP server that lets external AI agents query raw account data (nexus status, filings, transactions) — e.g. 'which states have I triggered nexus' — but this is infrastructure for connecting third-party/generic AI agents, not a first-party built-in tax-Q&A assistant that answers 'is this taxable' or 'why did this invoice tax this way' from its own rule base with citations. No evidence of a native chatbot or rule-cited answer engine exists in the pack.", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-docs-6" ] }, { "productId": "numeral", "storyId": "api-interactive-docs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Docs describe API endpoints and objects (products, transactions, certificates) but there is no evidence of an interactive reference with runnable/try-it examples; the openapi probe found no spec at standard locations, suggesting no Swagger/Postman-style interactive console.", "evidenceIds": [ "numeral-docs-2", "numeral-docs-3", "numeral-probe-2" ] }, { "productId": "numeral", "storyId": "api-machine-spec", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Probes explicitly show all common OpenAPI/swagger spec paths return 404, and no evidence pack item links to a downloadable machine-readable spec; only human-readable docs and an llms.txt are found.", "evidenceIds": [ "numeral-probe-2", "numeral-probe-1" ] }, { "productId": "numeral", "storyId": "api-sandbox", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence of a sandbox/test-mode environment or test API keys distinct from production; docs cover product/tax objects, MCP server, and integrations but never mention sandbox or test-vs-live data separation.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "api-versioning-policy", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Numeral's docs show clear API versioning via dated version paths (e.g., v2026-03-01) in multiple endpoints, indicating a versioning scheme, but no evidence describes a documented deprecation policy, sunset timeline, or migration guidance for older versions. Missing for 10: explicit deprecation policy documentation, version sunset/migration guides, changelog of breaking changes.", "evidenceIds": [ "numeral-docs-3", "numeral-docs-4" ] }, { "productId": "numeral", "storyId": "autofiling-remittance", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Numeral's marketing claims automatic registration, filing, and remittance across jurisdictions 'on time, every time' and handles related mail from tax authorities, showing strong first-party support for the filing/remittance workflow. However, the evidence pack contains no mention of a specific late-filing penalty guarantee or indemnification, which is a core part of the story. Missing for 10: explicit vendor guarantee/indemnification language covering penalties for late filings, independent verification of on-time filing performance.", "evidenceIds": [ "numeral-docs-11", "numeral-docs-12", "numeral-docs-9", "numeral-docs-10" ] }, { "productId": "numeral", "storyId": "automation-bulk-operations", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence only shows single-item API operations (create one product, one transaction, one certificate_request) with no documented batch/bulk endpoints for processing many items in one call. Bulk operations across many items is a plausible axis for a tax compliance API but no evidence supports it.", "evidenceIds": [ "numeral-docs-2", "numeral-docs-3", "numeral-docs-5" ] }, { "productId": "numeral", "storyId": "automation-rules-engine", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows built-in system behaviors (nexus threshold triggering, certificate expiration tracking, automatic validation) but no capability for an AI-native user to author custom rules that trigger arbitrary actions on events — no rules engine, webhook/trigger configuration, or conditional automation framework is documented.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "automation-scheduled-jobs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Numeral's docs describe automated recurring compliance processes (auto-filing, remittance, certificate renewal tracking) as built-in product behavior, but there's no evidence that an AI-native user can programmatically schedule or configure recurring jobs/workflows via the API or MCP server — the MCP tools expose read/trigger actions (nexus status, certificate requests) rather than a scheduling primitive. missing for 10: any API/MCP capability to create, list, or manage recurring/scheduled jobs or workflows initiated by the user or agent.", "evidenceIds": [ "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-docs-11", "numeral-docs-13" ] }, { "productId": "numeral", "storyId": "automation-versioned-workflows", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of versioning, review, or rollback of automations/workflows in Numeral's docs; the product focuses on tax calculation, filing, and MCP-based querying, with no mention of automation history, diffing, or revert capabilities.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "billing-stack-integrations", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Numeral claims 'native integrations for the platforms your business already runs on' and supports connecting sales channels for nexus studies, indicating maintained connections to commerce/billing platforms feed data in without custom plumbing. However, evidence lacks specifics on which billing systems, payment processors, or commerce platforms are actually supported, how they're maintained, or independent confirmation of integration depth/reliability. Missing for 10: named list of specific integrations (Stripe, Shopify, etc.), documentation of integration setup/maintenance process, and independent/hands-on verification that these integrations work reliably.", "evidenceIds": [ "numeral-docs-8", "numeral-docs-14" ] }, { "productId": "numeral", "storyId": "calculation-accuracy-guarantee", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence of any accuracy guarantee, indemnification, warranty, or audit-support commitment for erroneous tax calculations—only feature descriptions of the API, MCP server, and pricing terms.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "compliant-invoice-fields", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Numeral's docs show tax calculation, transaction/calculation objects, VAT ID validation and nexus/filing management, but there is no evidence of the product generating or feeding compliant invoice documents themselves — no mention of VAT registration numbers on invoices, required invoice line-item fields, or locale-specific invoice formatting rules flowing to an invoicing system.", "evidenceIds": [ "numeral-docs-1", "numeral-docs-3", "numeral-docs-16", "numeral-docs-11" ] }, { "productId": "numeral", "storyId": "exempt-customer-api", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Numeral's docs confirm an `exemption-certificates` object and `certificate_request` workflow (trigger/read/cancel) plus certificate validation and expiration tracking, showing exemption handling exists at the API level. However, there's no evidence of explicitly marking a customer or transaction exempt with a specific exemption reason field, nor documentation showing that reason flows through tax calculation, reporting, and the filed return itself. Missing for 10: explicit API fields/endpoint for setting exemption reason on customer/transaction, documentation of reason propagation into calculation logic, and confirmation it appears in reporting/filed returns.", "evidenceIds": [ "numeral-docs-5", "numeral-docs-13" ] }, { "productId": "numeral", "storyId": "exemption-certificate-management", "verdict": "partial", "quality": 7, "confidence": "medium", "rationale": "Numeral's docs describe a full certificate_request object lifecycle (trigger, read state, cancel) and pricing page confirms automatic validation of uploaded certificates plus expiration tracking, renewals, and compliance monitoring, covering collection, validation, and storage/expiry well. However, there is no explicit evidence that certificates are automatically applied at tax calculation time to exempt a transaction. Missing for 10: direct documentation of certificate application/exemption logic during the calculation API call, and independent/hands-on confirmation of the end-to-end flow.", "evidenceIds": [ "numeral-docs-5", "numeral-docs-13" ] }, { "productId": "numeral", "storyId": "filing-calendar-visibility", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Numeral's docs confirm it manages filings and remittance end-to-end (\"Register, file, and remit... on time, every time\", MCP tools surfacing filings and nexus status) but there is no evidence of a calendar-style upcoming-returns view, no mention of remitted amounts display, and no mention of downloadable confirmations for submitted returns. missing for 10: calendar view of upcoming filings, remitted-amount reporting per filing, downloadable filing confirmations.", "evidenceIds": [ "numeral-docs-11", "numeral-docs-9", "numeral-docs-10" ] }, { "productId": "numeral", "storyId": "global-vat-gst", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Numeral markets itself as covering 'US and global jurisdictions' and offers live VAT ID validation, suggesting some non-US tax support, but there is no evidence of specific OSS/IOSS registration, UK/Canada/Australia threshold tracking, or global filing workflows described in the docs. Missing for 10: explicit OSS/IOSS scheme support, per-country (UK, Canada, Australia) registration threshold tracking, and documented global filing/remittance processes beyond the generic 'global jurisdictions' claim.", "evidenceIds": [ "numeral-docs-11", "numeral-docs-16" ] }, { "productId": "numeral", "storyId": "historical-exposure-analysis", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Numeral offers a nexus study and MCP tools that report current nexus position per jurisdiction (thresholds, sales/transaction counts), which partially supports exposure monitoring, but there is no evidence of importing historical transactions to back-calculate liability, or of penalty/interest estimation per jurisdiction. missing for 10: historical transaction import for backdated analysis, estimated liability calculation, penalty and interest estimation per jurisdiction.", "evidenceIds": [ "numeral-docs-9", "numeral-docs-10", "numeral-docs-14" ] }, { "productId": "numeral", "storyId": "liability-reporting", "verdict": "partial", "quality": 4, "confidence": "medium", "rationale": "Numeral tracks nexus position and thresholds per jurisdiction and manages filings/registrations, and its API/MCP exposes transactions, sales, and calculations that could underpin liability tracking, but there is no explicit documentation of a 'collected vs owed' liability view or accountant/GL-ready exports. missing for 10: explicit accrued liability by jurisdiction report, collected-vs-owed breakdown, documented GL/accountant export formats (e.g., CSV/QBO sync).", "evidenceIds": [ "numeral-docs-10", "numeral-docs-11", "numeral-docs-9", "numeral-docs-7" ] }, { "productId": "numeral", "storyId": "marketplace-facilitator-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack details Numeral's nexus tracking, tax calculation, transactions, and filings features, but nowhere mentions marketplace facilitator sales (e.g., Amazon, app stores) or the specific handling needed to track such sales for nexus purposes without double-remitting tax already collected by the facilitator. This is a real and applicable capability for sales tax compliance software, but no evidence addresses it.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "nexus-threshold-monitoring", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Numeral explicitly tracks nexus position per jurisdiction with live economic thresholds and current sales/transaction counts (numeral-docs-10), offers a free nexus study (numeral-docs-14), and surfaces nexus status via MCP queries (numeral-docs-9). However, there is no explicit evidence of proactive threshold-crossing alerts/notifications sent before a limit is hit — the docs describe on-demand querying of current status rather than an automated pre-crossing warning system. Missing for 10: documented proactive alert/notification mechanism triggered before threshold breach, evidence of alert delivery channels (email/dashboard push), and any independent confirmation of alert timing accuracy.", "evidenceIds": [ "numeral-docs-10", "numeral-docs-9", "numeral-docs-14", "numeral-docs-7" ] }, { "productId": "numeral", "storyId": "omnichannel-import", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "Numeral documents native integrations for existing sales/business platforms and an API for creating transactions/calculations, suggesting multi-channel ingestion, but the evidence never specifies which connectors exist (e.g., specific processors/storefronts) or confirms historical backfill/import of past transactions rather than just live transaction creation. Missing for 10: named connector list, explicit backfill/historical-import capability, and confirmation multiple channels roll into a unified filing dataset.", "evidenceIds": [ "numeral-docs-8", "numeral-docs-3", "numeral-docs-14" ] }, { "productId": "numeral", "storyId": "openness-api-parity", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Numeral's docs show broad API coverage of core dashboard functions—products, transactions, calculations, exemption certificates, nexus, filings, registrations—and the MCP server exposes these same objects as agent-callable tools, suggesting strong parity between UI and API/agent access. However, there is no explicit statement that 100% of UI functionality is API-accessible, and a probe for a canonical OpenAPI spec returned 404s, leaving completeness unverified. Missing for 10: an explicit UI-to-API feature parity statement, a confirmed complete OpenAPI/API reference, and independent verification that all dashboard actions (e.g., billing, certificate uploads) are API-reachable.", "evidenceIds": [ "numeral-docs-2", "numeral-docs-3", "numeral-docs-5", "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-probe-2" ] }, { "productId": "numeral", "storyId": "openness-full-export", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence of any data export feature or open-format data portability for user data (transactions, filings, certificates, etc.); evidence covers API usage, MCP tools, and integrations but nothing about exporting/leaving with data.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "openness-open-license", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Numeral is a closed SaaS tax-compliance product/API, not open-source software; source-code openness is a category error for this type of product.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "openness-self-host", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Numeral is a hosted sales-tax compliance SaaS (calculations, filings, registrations, MCP/API access), not a product designed to be self-hosted; the axis is a category error for this kind of managed compliance service.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "privacy-data-residency", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Numeral is a sales-tax compliance/API product, not infrastructure with data residency controls; no evidence pack content addresses region/residency choice, and this is a category mismatch rather than a gap in an applicable axis.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "privacy-no-training", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Numeral is a sales tax compliance/filing product, not an AI model provider or platform with a data-training policy relevant to this axis; no evidence pack content addresses AI training data usage at all.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "privacy-retention-controls", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "Numeral is a sales-tax compliance platform whose evidence pack covers tax calculation, filings, MCP access, and integrations, but contains no mention of data retention policies, deletion controls, or privacy/data-lifecycle settings that an AI-native user could configure.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "privacy-telemetry-optout", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence in the pack addresses telemetry, usage tracking, or an opt-out mechanism for Numeral's product or its AI/MCP features.", "evidenceIds": [] }, { "productId": "numeral", "storyId": "product-tax-codes", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Numeral's docs confirm a product-taxability model (creating `product` objects via POST /tax/products and computing tax based on address + product set), implying category-based tax logic, but the evidence never surfaces the actual taxonomy (SaaS, digital goods, services, physical goods) or a way for developers to map their catalog into it. missing for 10: explicit tax category list/taxonomy, documented mapping workflow from SKU to category, jurisdiction-specific category rules, and independent verification of correctness.", "evidenceIds": [ "numeral-docs-1", "numeral-docs-2" ] }, { "productId": "numeral", "storyId": "rate-rule-currency", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "Numeral's core pitch is that it handles tax calculation and returns accurate rates automatically via API (numeral-docs-1), and it manages filings/registrations across jurisdictions (numeral-docs-11), implying rate/rule maintenance is vendor-managed rather than requiring customer redeployment. However, there is no explicit documentation describing how rate/rule changes are tracked, versioned, or communicated to customers, nor any changelog or notice mechanism for jurisdiction rule updates. missing for 10: explicit documentation of automatic rate/rule update process, a changelog or notification system for jurisdiction changes, and confirmation that no client-side redeploy/config change is ever needed when rules change.", "evidenceIds": [ "numeral-docs-1", "numeral-docs-11" ] }, { "productId": "numeral", "storyId": "realtime-tax-calculation", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Docs confirm an API-based flow where you submit an address and products and get back an accurate tax amount, with a create-calculation step before transactions, supporting real-time per-address tax logic (numeral-docs-1, numeral-docs-3, numeral-docs-4). However, there's no explicit claim of rooftop-level (vs ZIP) precision or checkout-speed/latency guarantees, and no independent benchmarking. missing for 10: explicit rooftop-level accuracy claims, latency/performance benchmarks for checkout use, independent verification of real-time speed.", "evidenceIds": [ "numeral-docs-1", "numeral-docs-3", "numeral-docs-4", "numeral-probe-1" ] }, { "productId": "numeral", "storyId": "refund-adjustment-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers tax calculation, transactions, exemption certificates, nexus tracking, and filing/remittance, but nowhere mentions refunds, credit notes, or negative-amount adjustments flowing through liability and reporting. This is a fair axis for a tax compliance product, but no documentation addresses it.", "evidenceIds": [ "numeral-docs-3", "numeral-docs-11" ] }, { "productId": "numeral", "storyId": "saas-digital-taxability", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows only generic tax calculation via address+product lookup (numeral-docs-1) and certificate/nexus tracking, but nothing about SaaS-specific state-by-state taxability nuances, digital-goods rules, bundled-offering treatment, or visible research/rationale behind determinations.", "evidenceIds": [ "numeral-docs-1", "numeral-docs-2" ] }, { "productId": "numeral", "storyId": "state-registrations", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Numeral explicitly claims to 'Register, file, and remit across US and global jurisdictions' on the customer's behalf, and its MCP/API surface exposes 'registrations' as an object with status data (nexus position, filing review status), supporting in-product tracking. However, there is no evidence of transparent per-registration pricing — the pricing docs describe billing tied to filings generally, not itemized registration costs. Missing for 10: explicit per-registration price disclosure, and independent/hands-on confirmation that registrations are actually filed (not just claimed).", "evidenceIds": [ "numeral-docs-11", "numeral-docs-7", "numeral-docs-9", "numeral-docs-10", "numeral-docs-15" ] }, { "productId": "numeral", "storyId": "transaction-audit-trail", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "Numeral's API models a `calculation` object tied to each `transaction`, uses address/IP for resolution, and ties products to tax categories, suggesting the underlying data needed for reconstruction exists (numeral-docs-1,2,3,4). However, there is no evidence of an audit-export feature, a persisted rate/rule breakdown, or a UI/API for retrieving historical calculation detail for audit purposes.\nmissing for 10: explicit audit-trail/export functionality, documentation of retained rate and rule detail per transaction, and any mention of long-term storage/retrieval of calculation inputs for compliance review.", "evidenceIds": [ "numeral-docs-1", "numeral-docs-2", "numeral-docs-3", "numeral-docs-4" ] }, { "productId": "numeral", "storyId": "vat-id-validation", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Numeral's pricing page claims 'Validate customer VAT IDs live' (numeral-docs-16), suggesting some real-time VAT ID validation capability, but there is no documentation of reverse charge mechanics, official registry (e.g., VIES) integration, or B2B cross-border transaction logic anywhere in the evidence pack — the product's core narrative is US sales tax nexus/filing, not EU VAT compliance. Missing for 10: reverse charge application logic, explicit official-registry validation (VIES or equivalent), B2B cross-border transaction handling, and any independent corroboration of the VAT ID validation claim.", "evidenceIds": [ "numeral-docs-16" ] }, { "productId": "numeral", "storyId": "vda-remediation", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence covers nexus tracking, registration, filing, remittance, and mail handling, but nothing mentions voluntary disclosure agreements (VDAs) or back-filing/back-tax remediation support for past-due exposure in jurisdictions where the company should have registered earlier.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "agent-reads-tax-state", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Stripe Tax exposes API access to registrations, calculations, filings, and reports (docs-7,8,10,11,14), and Stripe has a documented MCP server (stripe-tax-probe-4), but there's no evidence the MCP surface exposes tax-specific read/act capabilities (nexus status, accrued liability, upcoming filings) or scoped credentials for an agent to autonomously act on tax posture. missing for 10: MCP server tool list showing tax-specific endpoints (nexus/liability/filings), scoped-credential/permission model for agentic access, and any hands-on/community confirmation of agent-driven tax actions via API or MCP.", "evidenceIds": [ "stripe-tax-docs-7", "stripe-tax-docs-8", "stripe-tax-docs-10", "stripe-tax-docs-11", "stripe-tax-docs-14", "stripe-tax-probe-4" ] }, { "productId": "stripe-tax", "storyId": "agentic-agent-docs", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Direct probe evidence confirms an llms.txt file exists at docs.stripe.com/llms.txt (HTTP 200) and per-page markdown docs (e.g. tax.md) are also served, allowing agents to fetch agent-oriented documentation directly. This is Stripe-wide infrastructure that applies to the Tax docs specifically (stripe-tax-docs-2 sourced from tax.md). Missing for 10: no evidence of a dedicated llms-full.txt or Tax-specific agent guide, and no independent/community confirmation of agents successfully using this in practice.", "evidenceIds": [ "stripe-tax-probe-1", "stripe-tax-probe-2", "stripe-tax-docs-2" ] }, { "productId": "stripe-tax", "storyId": "agentic-ai-insights", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "The axis applies to this product kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is \"none\", never \"na\". (na/none harmonized at arena bring-up — see pipeline/scripts/na-harmonize.ts.)", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "agentic-autonomous-automation", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Stripe Tax's core value is automatic tax calculation on every transaction (docs-2, comm-1) and automated US filing via TaxJar integration (docs-10, comm-5), which are background automations that run without manual intervention. However there is no evidence of an AI-native workflow where a user configures agentic automations (e.g., via the general Stripe MCP server or CLI) specifically for tax tasks running autonomously. Missing for 10: explicit AI-agent-configurable automation/trigger setup for tax workflows, evidence of the MCP/CLI being used to orchestrate autonomous background tax processes, and independent confirmation that these automations run without ongoing manual oversight (comm-4 notes filing still requires manual paperwork/payment).", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-10", "stripe-tax-comm-1", "stripe-tax-comm-5", "stripe-tax-comm-4", "stripe-tax-probe-4" ] }, { "productId": "stripe-tax", "storyId": "agentic-builtin-assistant", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "The axis applies to this product kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is \"none\", never \"na\". (na/none harmonized at arena bring-up — see pipeline/scripts/na-harmonize.ts.)", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "agentic-headless", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Stripe Tax is fully API-driven (tax calculation, sandbox testing, itemized exports) and Stripe provides a CLI and sandbox environment usable in CI/automation pipelines, supporting headless operation. However, the evidence is generic Stripe platform tooling rather than Stripe Tax-specific CI/automation documentation, and no CI examples, GitHub Actions, or automated testing pipeline integration specific to Tax are shown. missing for 10: Stripe Tax-specific CI/CD pipeline examples, automated testing workflows referencing Tax API, evidence of headless batch/report automation beyond manual sandbox testing.", "evidenceIds": [ "stripe-tax-docs-6", "stripe-tax-docs-13", "stripe-tax-probe-5", "stripe-tax-docs-11" ] }, { "productId": "stripe-tax", "storyId": "agentic-mcp-client", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax-calculation API/SaaS product, not an agentic system that itself consumes external tools via MCP; the only MCP-related evidence shows Stripe exposes an MCP server so *other* agents can call Stripe's tools, which is the reverse relationship and irrelevant to this story about the product acting as an MCP client.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "agentic-mcp-server", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Stripe (the platform Stripe Tax belongs to) documents an official MCP server at docs.stripe.com/mcp, evidencing agent connectivity, but the evidence pack has no Stripe Tax-specific MCP tooling details, hands-on agent usage, or independent corroboration of MCP functioning with Stripe Tax specifically. Missing for 10: Stripe Tax-specific MCP tool documentation, independent/hands-on validation of the MCP server working with tax operations, details on scope/coverage of tax-related MCP tools.", "evidenceIds": [ "stripe-tax-probe-4" ] }, { "productId": "stripe-tax", "storyId": "agentic-nl-commands", "verdict": "partial", "quality": 5, "confidence": "low", "rationale": "Stripe (the platform) documents an official MCP server and CLI (stripe-tax-probe-4, stripe-tax-probe-5) plus an llms.txt for docs discovery, which are the building blocks for AI-agent/natural-language operation, but none of the evidence shows these specifically exposing Stripe Tax's calculation/filing/registration workflows via natural-language commands. Missing for 10: tax-specific MCP tool examples, natural-language usage documentation for Tax operations, and independent confirmation of AI-agent control over tax settings.", "evidenceIds": [ "stripe-tax-probe-4", "stripe-tax-probe-5", "stripe-tax-probe-1" ] }, { "productId": "stripe-tax", "storyId": "agentic-official-cli", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "Stripe (the parent platform) documents an official CLI at docs.stripe.com/stripe-cli, which covers Stripe Tax API usage as part of the broader Stripe API surface. Missing for 10: no Stripe Tax-specific CLI commands or documentation, and no independent/community corroboration of CLI usage specifically for tax workflows.", "evidenceIds": [ "stripe-tax-probe-5" ] }, { "productId": "stripe-tax", "storyId": "agentic-public-api", "verdict": "full", "quality": 8, "confidence": "high", "rationale": "Stripe Tax is fully API-driven, with documented parameters (shipping_cost, IP geolocation, ship-from address), a sandbox/testing environment, and taxability_reason fields returned via API responses, plus a standalone Tax API doc. Community evidence confirms real-world programmatic integration (VAT/tax rate automation) though with operational gotchas unrelated to API documentation quality. Missing for 10: a discoverable OpenAPI spec (probe returned 404s) and independent hands-on confirmation of API robustness beyond docs.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-3", "stripe-tax-docs-4", "stripe-tax-docs-5", "stripe-tax-docs-6", "stripe-tax-docs-13", "stripe-tax-probe-2", "stripe-tax-probe-3", "stripe-tax-comm-1" ] }, { "productId": "stripe-tax", "storyId": "agentic-scoped-keys", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence of Stripe Tax (or Stripe's platform) offering scoped/least-privilege API credential issuance for agents; the pack discusses general API access (restricted keys) nowhere, only tax calculation features, CLI, and MCP server pointers. Missing for 10: any mention of restricted/scoped API keys, OAuth scopes, or agent-specific credential controls.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "agentic-sdks", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Evidence shows Stripe's docs reference official npm packages/SDKs for integration (stripe-tax-probe-1) and Stripe Tax is built on the standard Stripe API with docs pages (stripe-tax-docs-1..14), implying SDK-based access, but the pack contains no explicit list of official language SDKs, code samples in multiple languages, or SDK-specific documentation for Tax endpoints. missing for 10: explicit official SDK list/pages for Stripe Tax, multi-language code samples, versioning/release notes for SDKs.", "evidenceIds": [ "stripe-tax-probe-1", "stripe-tax-docs-2", "stripe-tax-docs-3" ] }, { "productId": "stripe-tax", "storyId": "agentic-webhooks", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "The evidence pack contains extensive Stripe Tax documentation on tax calculation, registration, filing, and reporting, but no mention of webhooks, event subscriptions, or any event-driven integration mechanism. Without evidence of webhook support for Stripe Tax events, this applicable axis cannot be credited.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "ai-taxability-answers", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Evidence shows Stripe Tax provides rule-based calculation, reporting, and a taxability_reason field for API results, but there is no evidence of a conversational AI/chatbot answering ad-hoc questions like 'is this taxable in this state' with citations. This is a rules/API engine, not an AI-agent Q&A interface. Missing for 10: any AI chat/assistant interface, natural-language Q&A capability, citation-backed reasoning beyond a single API field.", "evidenceIds": [ "stripe-tax-docs-13", "stripe-tax-docs-2" ] }, { "productId": "stripe-tax", "storyId": "api-interactive-docs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack includes Stripe Tax conceptual docs, an MCP server, and a CLI, but nothing describes an interactive API reference with runnable/try-it code examples; the openapi probe even returned 404s for all candidate spec locations, suggesting no discoverable machine-readable API spec for this evidence pack.", "evidenceIds": [ "stripe-tax-probe-3" ] }, { "productId": "stripe-tax", "storyId": "api-machine-spec", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack includes an explicit probe for OpenAPI/Swagger spec files at all common locations, all returning 404, and no other citation points to a downloadable machine-readable API spec for Stripe Tax; only human-readable docs and an MCP/CLI mention exist.", "evidenceIds": [ "stripe-tax-probe-3" ] }, { "productId": "stripe-tax", "storyId": "api-sandbox", "verdict": "full", "quality": 8, "confidence": "medium", "rationale": "Docs explicitly confirm a sandbox environment can be used to preview automatic tax calculation for different tax settings without touching production, and Stripe's broader test-mode/sandbox architecture is documented via CLI/API tooling. Missing for 10: independent hands-on confirmation of sandbox parity with production data isolation, and no explicit mention of test API keys/test clocks specific to Tax.", "evidenceIds": [ "stripe-tax-docs-6", "stripe-tax-probe-5" ] }, { "productId": "stripe-tax", "storyId": "api-versioning-policy", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence in the pack discusses API versioning strategy or a documented deprecation policy for Stripe Tax's API; docs cover tax calculation features, CLI/MCP tooling, and community pricing/reliability complaints, none of which address versioning or deprecation commitments.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "autofiling-remittance", "verdict": "disputed", "quality": 2, "confidence": "medium", "rationale": "Stripe Tax only offers automated filing via a TaxJar integration for US sales tax (stripe-tax-docs-10), not for 'every registered jurisdiction,' and there is no evidence of any penalty guarantee for late filing. Community evidence directly contradicts the 'automatic filing/remittance' claim: users report 'you still have to do everything yourself - filing the papers and wiring money' and recommend using a Merchant of Record instead specifically to avoid doing filing/remittance manually. missing for 10: global (non-US) automated filing coverage, any late-filing penalty guarantee, evidence that remittance itself is automated rather than manual.", "evidenceIds": [ "stripe-tax-docs-10", "stripe-tax-comm-4", "stripe-tax-comm-5", "stripe-tax-comm-11" ] }, { "productId": "stripe-tax", "storyId": "automation-bulk-operations", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Stripe Tax docs describe per-transaction tax calculation (including multi-line-item support via shipping_cost) and itemized exports/reports, but there is no evidence of a batch/bulk API endpoint or workflow that lets an AI-native user submit or calculate tax across many separate transactions/items in a single bulk operation. missing for 10: bulk/batch tax-calculation API, documented batch endpoints, evidence of processing many orders in one automated call.", "evidenceIds": [ "stripe-tax-docs-3", "stripe-tax-docs-11", "stripe-tax-docs-2" ] }, { "productId": "stripe-tax", "storyId": "automation-rules-engine", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax calculation/filing service, not a rules-engine/automation platform for defining event-triggered actions; this axis is a category error for this product type.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "automation-scheduled-jobs", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax calculation/compliance API, not a workflow/job-scheduling platform; scheduling recurring automation jobs is outside its product category.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "automation-versioned-workflows", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax calculation/compliance service, not an automation-building or workflow-authoring tool; versioning, reviewing, and rolling back automations is not an applicable axis for this product category.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "billing-stack-integrations", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Stripe Tax is deeply integrated with Stripe's own revenue products (Billing, Invoicing, Checkout, Payment Links) and offers a standalone Tax API that can calculate tax even on payments processed outside Stripe, plus a maintained TaxJar integration for automated US filing. However, there is no evidence of maintained, pre-built connectors to third-party billing systems, payment processors, or commerce platforms (e.g., Shopify, WooCommerce, other PSPs) — using it outside Stripe's own ecosystem still requires developers to build the API plumbing themselves, and community feedback notes it is 'woefully incomplete' for international sellers and inferior to dedicated tax platforms like Avalara that have broader integration ecosystems. missing for 10: documented third-party commerce/payment-processor integrations beyond Stripe's own products and TaxJar, independent confirmation these integrations are actively maintained and cover non-Stripe stacks.", "evidenceIds": [ "stripe-tax-docs-1", "stripe-tax-docs-10", "stripe-tax-comm-5", "stripe-tax-comm-9", "stripe-tax-comm-10" ] }, { "productId": "stripe-tax", "storyId": "calculation-accuracy-guarantee", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence pack item mentions an accuracy guarantee, indemnification, or audit support commitment from Stripe if a tax calculation is wrong; documentation covers calculation mechanics, filing, and reporting only, and community threads discuss billing bugs and pricing complaints rather than any guarantee/warranty language. missing for 10: any documented accuracy guarantee, liability/indemnification clause, or audit-defense commitment.", "evidenceIds": [ "stripe-tax-docs-1", "stripe-tax-docs-2", "stripe-tax-comm-6", "stripe-tax-comm-7" ] }, { "productId": "stripe-tax", "storyId": "compliant-invoice-fields", "verdict": "disputed", "quality": 5, "confidence": "medium", "rationale": "Stripe Tax calculates VAT/GST by locale, computes rates based on customer/business location, integrates with Invoicing/Checkout, and supports registration number handling and tax-code based rules (stripe-tax-docs-2,4,5,8,12,14), with a positive first-party account of it automating VAT rate/number syncing (stripe-tax-comm-1). However, concrete community evidence contradicts full 'compliant everywhere' delivery: VIES VAT-number validation only fires once with no re-trigger, requiring manual quarterly re-checks for compliance (stripe-tax-comm-2), a documented incident where Stripe Tax silently stopped billing many customers and left invoices stuck in Draft (stripe-tax-comm-6, acknowledged by a Stripe employee in stripe-tax-comm-7), and reports that international coverage is 'woefully incomplete' (stripe-tax-comm-10). Missing for 10: explicit documentation of which specific VAT fields/registration numbers are auto-populated on the invoice template itself, and independent confirmation that the Draft-invoice/VIES issues have been fully remediated.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-4", "stripe-tax-docs-5", "stripe-tax-docs-8", "stripe-tax-docs-12", "stripe-tax-docs-14", "stripe-tax-comm-1", "stripe-tax-comm-2", "stripe-tax-comm-6", "stripe-tax-comm-7", "stripe-tax-comm-10" ] }, { "productId": "stripe-tax", "storyId": "exempt-customer-api", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows Stripe Tax can return a taxability_reason for a calculated tax result and supports itemized reporting, but nowhere does it document an API mechanism to mark a customer or transaction as tax-exempt with a specified exemption reason, nor any linkage of that reason through to the filed return. Missing for 10: exemption-marking API/parameter, exemption certificate storage, and evidence the exemption reason propagates into reports and filed returns.", "evidenceIds": [ "stripe-tax-docs-13", "stripe-tax-docs-11", "stripe-tax-docs-10" ] }, { "productId": "stripe-tax", "storyId": "exemption-certificate-management", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence in the pack describes collection, validation, storage, expiry tracking, or automatic application of exemption/resale certificates; the closest related item is VAT number validation (VIES), which is a different mechanism and explicitly noted as one-time only with no re-validation. This is a plausible axis for a tax compliance product, but nothing here shows Stripe Tax supports it end to end.", "evidenceIds": [ "stripe-tax-comm-2" ] }, { "productId": "stripe-tax", "storyId": "filing-calendar-visibility", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence shows Stripe Tax offers filing automation via TaxJar integration (stripe-tax-docs-10) and itemized tax exports (stripe-tax-docs-11), but there is no mention of a filing calendar, per-filing status tracker, remitted-amount summaries, or downloadable filing confirmations. A community report even states that despite Stripe's tooling users 'still have to do everything yourself - filing the papers and wiring money' (stripe-tax-comm-4), reinforcing the absence of a consolidated filing dashboard. Missing for 10: filing calendar view, per-filing status indicators, remitted amount tracking, downloadable confirmation documents.", "evidenceIds": [ "stripe-tax-docs-10", "stripe-tax-docs-11", "stripe-tax-comm-4" ] }, { "productId": "stripe-tax", "storyId": "global-vat-gst", "verdict": "disputed", "quality": 4, "confidence": "medium", "rationale": "Stripe Tax's docs confirm VAT/GST calculation across EU, UK, Canada, Australia and registration-threshold monitoring (docs-2, docs-7, docs-8, docs-14), but the filing-support integration (TaxJar) is explicitly scoped to 'US sales tax' only (stripe-tax-docs-10), and community evidence directly contradicts broader filing/remittance support — one user notes 'you still have to do everything yourself - filing the papers and wiring money' and another says 'If you're selling internationally it's woefully incomplete, so often not that useful' (stripe-tax-comm-4, stripe-tax-comm-10), with a third suggesting a Merchant of Record instead for EU/UK remittance (stripe-tax-comm-11). missing for 10: automated filing/remittance support for VAT/GST jurisdictions beyond US sales tax, and evidence rebutting the community reports of incomplete international coverage.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-7", "stripe-tax-docs-8", "stripe-tax-docs-10", "stripe-tax-comm-4", "stripe-tax-comm-10", "stripe-tax-comm-11" ] }, { "productId": "stripe-tax", "storyId": "historical-exposure-analysis", "verdict": "partial", "quality": 3, "confidence": "medium", "rationale": "Stripe Tax docs mention identifying where you 'might need to register to collect tax based on past transactions' (nexus monitoring), but there is no evidence of importing historical transactions to produce an estimated liability, penalties, and interest per jurisdiction — the core of this backdated exposure-analysis story. missing for 10: historical transaction import, estimated back-tax liability calculation, penalty/interest estimation per jurisdiction, any hands-on or documented feature matching this workflow.", "evidenceIds": [ "stripe-tax-docs-7", "stripe-tax-docs-8" ] }, { "productId": "stripe-tax", "storyId": "liability-reporting", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Stripe Tax provides itemized tax reports/exports and TaxJar-based US filing assistance, giving some filing-ready detail, but there is no documented dashboard showing accrued liability by jurisdiction (collected vs. owed) or GL-mapped exports for accountants, and community reports note filing/remittance still requires manual work and international coverage gaps. missing for 10: explicit collected-vs-owed liability view, GL/accounting-system export format, evidence of jurisdiction-level real-time liability tracking.", "evidenceIds": [ "stripe-tax-docs-11", "stripe-tax-docs-10", "stripe-tax-comm-4", "stripe-tax-comm-10" ] }, { "productId": "stripe-tax", "storyId": "marketplace-facilitator-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence in the pack addresses marketplace-facilitator tax collection (Amazon, app stores) or how Stripe Tax tracks facilitator-collected transactions for nexus/reporting purposes without double-remitting. All docs and community items focus on Stripe Tax's own calculation/filing for payments processed through Stripe, not on handling third-party facilitator-collected sales.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "nexus-threshold-monitoring", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Stripe Tax docs mention identifying registration obligations based on past transactions and registering 'as soon as you start processing payments,' which implies some threshold-tracking capability, but there is no evidence of proactive, pre-crossing alerts or real-time dashboards tracking sales/transaction counts against every state and country threshold before they're exceeded. Community evidence focuses on calculation and filing integration, not nexus threshold alerting. Missing for 10: explicit alerting/notification mechanism before threshold crossing, real-time nexus dashboard evidence, and independent confirmation that monitoring is forward-looking rather than retrospective.", "evidenceIds": [ "stripe-tax-docs-7", "stripe-tax-docs-8", "stripe-tax-docs-1" ] }, { "productId": "stripe-tax", "storyId": "omnichannel-import", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "Stripe Tax docs claim it can 'calculate and report tax on all payments, even those you process outside of Stripe,' suggesting some cross-channel coverage, but there is no evidence of concrete connector integrations with other processors/storefronts or of an API-based historical backfill mechanism for importing past transactions from other channels. Missing for 10: documented connectors/integrations for third-party processors or storefronts, API endpoints for bulk/backfill import of historical transactions, and independent confirmation these actually cover 'whole business' filings.", "evidenceIds": [ "stripe-tax-docs-1", "stripe-tax-probe-2" ] }, { "productId": "stripe-tax", "storyId": "openness-api-parity", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Docs show deep API coverage for tax calculation (shipping cost, IP geolocation, ship-from address, sandbox testing, itemized reports, taxability_reason) mirroring UI capabilities, and a documented CLI/MCP exist for API-driven workflows. However, registration and filing—explicitly UI-flagged features (docs-7, docs-8, docs-14, docs-10)—appear to rely on Stripe performing them 'on your behalf' or via TaxJar integration rather than exposing full programmatic control, and no evidence confirms these can be fully driven via API alone. missing for 10: explicit API endpoints/parity for tax registration and filing workflows, independent confirmation of full UI-API parity.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-3", "stripe-tax-docs-4", "stripe-tax-docs-5", "stripe-tax-docs-6", "stripe-tax-docs-11", "stripe-tax-docs-13", "stripe-tax-docs-7", "stripe-tax-docs-8", "stripe-tax-docs-14", "stripe-tax-docs-10", "stripe-tax-probe-5", "stripe-tax-probe-4" ] }, { "productId": "stripe-tax", "storyId": "openness-full-export", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Stripe Tax offers itemized exports with line-item tax breakdowns per transaction, which is a data-export capability, but there is no evidence of a comprehensive 'export all data' feature, explicit open-format guarantees (e.g., CSV/JSON specification), or support for fully leaving the platform with your data. missing for 10: full data export scope beyond tax reports, explicit open-format documentation, independent confirmation of unrestricted data portability.", "evidenceIds": [ "stripe-tax-docs-11" ] }, { "productId": "stripe-tax", "storyId": "openness-open-license", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a closed-source commercial SaaS tax-calculation service, not an open-source project; source availability under an open license is not a plausible axis for this kind of hosted product.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "openness-self-host", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a hosted SaaS tax-calculation service; self-hosting the core product is a category error for this kind of managed cloud service, not an applicable capability.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "privacy-data-residency", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax calculation/compliance product, not a data-storage/hosting platform; data residency/region selection is not a relevant capability axis for this product type.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "privacy-no-training", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax calculation/compliance product, not an AI model or AI-native tool with training-data policies; the question of preventing data from being used for AI training is a category error for this axis.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "privacy-retention-controls", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax calculation/reporting product, not an AI agent or data-processing platform with configurable data retention/deletion controls; this privacy-posture axis about controlling AI data retention is a category error for this product type.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "privacy-telemetry-optout", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "Stripe Tax is a tax-calculation service, not an AI agent or SDK with telemetry collection; opting out of usage tracking is not a relevant axis for this product type, and no evidence discusses telemetry.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "product-tax-codes", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Stripe Tax documents product tax codes (PTCs) that determine tax rates based on product type (SaaS, digital goods, services, physical goods) and lets Stripe apply a preset PTC when none is set on a specific item, which is exactly the catalog-mapping mechanism the story asks for. However, community feedback specifically flags coverage gaps when selling internationally ('woefully incomplete' for international jurisdictions), and there's no independent verification that every category/jurisdiction combination is accurately handled. Missing for 10: independent/hands-on confirmation of correct taxability across all four category types and all jurisdictions, and clearer documentation addressing the international completeness gap raised by users.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-12", "stripe-tax-docs-13", "stripe-tax-comm-10" ] }, { "productId": "stripe-tax", "storyId": "rate-rule-currency", "verdict": "partial", "quality": 7, "confidence": "medium", "rationale": "Docs and community evidence confirm Stripe Tax automatically calculates rates across jurisdictions worldwide based on location/product type without manual rate syncing, and a merchant explicitly notes it eliminated their need to manually sync VAT rates. However, the pack has no evidence of a changelog or documentation specifying exactly when rate/rule changes take effect, and no explicit statement guaranteeing zero-redeploy behavior beyond the general 'automatic' framing. Missing for 10: a public changelog/versioning of rate updates, explicit no-redeploy guarantee, and independent verification that rule updates propagate without any integration changes.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-5", "stripe-tax-comm-1" ] }, { "productId": "stripe-tax", "storyId": "realtime-tax-calculation", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Stripe Tax docs confirm real-time API tax calculation based on full customer address, business location, product type, and even IP-based geolocation before an address is captured, which is well beyond ZIP-code-level approximation and is used inside checkout/payment flows (stripe-tax-docs-2, stripe-tax-docs-4, stripe-tax-docs-5, stripe-tax-comm-1). However, there's no explicit documentation of latency/performance guarantees for checkout-speed use, no explicit 'rooftop-accuracy' claim, and community feedback flags that international coverage is 'woefully incomplete' compared to competitors like Avalara (stripe-tax-comm-9, stripe-tax-comm-10), tempering confidence in universal address-level precision. Missing for 10: explicit rooftop/address-level accuracy documentation, published latency/performance benchmarks for checkout integration, and independent verification of accuracy claims across all jurisdictions.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-docs-4", "stripe-tax-docs-5", "stripe-tax-comm-1", "stripe-tax-comm-9", "stripe-tax-comm-10" ] }, { "productId": "stripe-tax", "storyId": "refund-adjustment-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack covers tax calculation, registration, filing integration with TaxJar, and reporting exports, but nothing specifically addresses how refunds or credit notes adjust tax liability/reporting/filing, nor confirms negative amounts flow through without manual journal fixes. This is a reasonable axis for a tax product to be judged on, but no supporting evidence is present.", "evidenceIds": [] }, { "productId": "stripe-tax", "storyId": "saas-digital-taxability", "verdict": "disputed", "quality": 4, "confidence": "medium", "rationale": "Stripe Tax docs show product tax codes (PTC) driving taxability determinations and a `taxability_reason` field exposing the rationale for a given tax result (docs-12, docs-13), which partially supports the 'visible research' ask, but there's no documentation specifically addressing state-by-state SaaS taxability nuances or bundled-offering treatment. Community evidence directly contradicts the completeness claim — practitioners call it 'vastly outgunned by its competitors' and 'woefully incomplete' for international/complex sales (comm-9, comm-10), undercutting confidence it reliably nails the hard cases. Missing for 10: documented state-specific SaaS/digital-goods taxability guidance, bundled-offering tax logic, and independent verification that edge-case determinations are correct.", "evidenceIds": [ "stripe-tax-docs-12", "stripe-tax-docs-13", "stripe-tax-comm-9", "stripe-tax-comm-10" ] }, { "productId": "stripe-tax", "storyId": "state-registrations", "verdict": "disputed", "quality": 4, "confidence": "medium", "rationale": "Stripe docs claim Stripe can 'register on your behalf' for tax permits and that Stripe Tax 'integrates with TaxJar to automate filing for US sales tax,' but a hands-on community account explicitly contradicts the filing automation, stating 'you still have to do everything yourself - filing the papers and wiring money' after using Stripe's tax tools. No evidence anywhere shows per-registration pricing transparency or a status-tracking UI for registrations. missing for 10: documented per-registration pricing, in-product registration status tracker, evidence that state/local filings (beyond TaxJar's limited US sales tax scope) are actually filed by Stripe rather than the merchant.", "evidenceIds": [ "stripe-tax-docs-8", "stripe-tax-docs-10", "stripe-tax-docs-14", "stripe-tax-comm-4", "stripe-tax-comm-5" ] }, { "productId": "stripe-tax", "storyId": "transaction-audit-trail", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "Docs show itemized exports with line-item tax breakdown, a taxability_reason field explaining the calculation logic, product tax codes, and address-resolution methods (IP geolocation, ship-from address), which together support reconstructing rate, category, and address logic per transaction. However there's no explicit documentation of data retention duration/policy, no audit-specific report bundling all these fields together, and no independent confirmation that exports satisfy audit requirements. Missing for 10: explicit retention/export lifecycle policy, a unified audit trail report combining rate+rules+address+category, and independent/hands-on verification of audit usability.", "evidenceIds": [ "stripe-tax-docs-11", "stripe-tax-docs-13", "stripe-tax-docs-12", "stripe-tax-docs-4", "stripe-tax-docs-5" ] }, { "productId": "stripe-tax", "storyId": "vat-id-validation", "verdict": "disputed", "quality": 4, "confidence": "medium", "rationale": "Stripe Tax docs confirm automated VAT/GST calculation based on customer/business location and tax codes, and a community report confirms VAT-number (VIES) validation exists, but that same first-hand account states validation only fires once and is never re-triggered — contradicting a true 'real-time' registry check — and no evidence anywhere documents explicit reverse-charge logic for B2B cross-border sales. Missing for 10: explicit reverse-charge documentation, continuous/real-time VIES re-validation, and independent confirmation of correct B2B cross-border handling beyond the disputed one-time check.", "evidenceIds": [ "stripe-tax-docs-2", "stripe-tax-comm-1", "stripe-tax-comm-2", "stripe-tax-comm-10" ] }, { "productId": "stripe-tax", "storyId": "vda-remediation", "verdict": "disputed", "quality": 4, "confidence": "medium", "rationale": "Stripe Tax docs claim to help identify where past transactions create registration exposure and can register on your behalf, but there's no evidence of voluntary disclosure agreement (VDA) support or back-filing assistance for cleaning up historical non-compliance. A community account (comm-4) directly contradicts any remediation narrative, stating that even with Stripe's new tools 'you still have to do everything yourself - filing the papers and wiring money,' undercutting the impression of an end-to-end remediation path. Missing for 10: explicit VDA/back-filing support, evidence of handling historical liability cleanup, and confirmation that registration assistance extends beyond forward-looking compliance.", "evidenceIds": [ "stripe-tax-docs-7", "stripe-tax-docs-14", "stripe-tax-comm-4" ] }, { "productId": "taxjar", "storyId": "agent-reads-tax-state", "verdict": "partial", "quality": 3, "confidence": "medium", "rationale": "TaxJar has a documented REST API with API-key auth (docs-1,2,4) and a nexus-threshold tracking feature (docs-7,8), but evidence shows no API/agent access to registrations or accrued liability, no scoped/limited credentials beyond a flat API key, and no MCP surface — probes for llms.txt and OpenAPI discovery both failed (taxjar-probe-1, taxjar-probe-2). Missing for 10: API endpoints for registrations/accrued liability/filings, scoped/least-privilege credential model, any MCP server or agent-facing surface, independent confirmation of nexus data via API rather than dashboard-only.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-2", "taxjar-docs-7", "taxjar-docs-8", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-agent-docs", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Direct probes show no llms.txt (404) and no discoverable OpenAPI spec, and no evidence of agent-oriented documentation exists anywhere in the pack.", "evidenceIds": [ "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-ai-insights", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of AI-generated insights, recommendations, or in-product analytics powered by AI; the nexus threshold tracker (taxjar-docs-7/8) is rule-based reporting, not AI-generated. Missing for 10: any mention of AI/ML-driven insights, natural-language summaries, or suggestion features within TaxJar.", "evidenceIds": [ "taxjar-docs-7", "taxjar-docs-8" ] }, { "productId": "taxjar", "storyId": "agentic-autonomous-automation", "verdict": "partial", "quality": 4, "confidence": "medium", "rationale": "TaxJar's AutoFile feature runs autonomously in the background to file and remit state tax returns without manual intervention, which is a genuine 'set-and-forget' automation. However, this is a single fixed built-in feature rather than a general automation framework letting AI-native users configure their own triggers/workflows, and there is no evidence of webhooks, scheduled jobs, or agent-friendly automation tooling beyond the core API. Missing for 10: general-purpose automation/trigger framework, webhook or event-driven automation support, evidence of AI-agent-configurable workflows beyond AutoFile.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8" ] }, { "productId": "taxjar", "storyId": "agentic-builtin-assistant", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "The axis applies to this product kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is \"none\", never \"na\". (na/none harmonized at arena bring-up — see pipeline/scripts/na-harmonize.ts.)", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "agentic-headless", "verdict": "full", "quality": 6, "confidence": "medium", "rationale": "TaxJar exposes a full REST API with API-key auth and official client libraries in seven languages, which supports headless/CI automation for tax calculations, transaction uploads, and nexus tracking. Missing for 10: no published OpenAPI/Swagger spec or llms.txt (both probes 404), no explicit CI/CLI examples or independent hands-on automation reports.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-2", "taxjar-docs-3", "taxjar-docs-4", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-mcp-client", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "TaxJar is a tax-calculation SaaS API/service, not an AI agent host that could consume MCP servers as tool providers; the evidence pack only covers TaxJar's own REST API, integrations, and clients, with no mention of MCP client functionality. This axis (plugging MCP servers into the product) is a category error for this kind of product.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "agentic-mcp-server", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "TaxJar is a SaaS tax API/service, so an official MCP server is a fair question, but evidence only shows a traditional REST API with language SDKs, no MCP server; probes for llms.txt and OpenAPI spec both returned 404s, indicating no AI-native integration surface.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-3", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-nl-commands", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of any natural-language interface, chatbot, or NL command layer; TaxJar exposes only a traditional REST API with SDKs, and probes for llms.txt/OpenAPI discovery artifacts returned 404s. Missing for 10: any NL command support, conversational interface, or agent-friendly API documentation.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-2", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-official-cli", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of an official CLI tool; TaxJar offers API clients in several languages but nothing resembling a command-line interface, and probes for openapi/llms.txt failed, suggesting no such tooling exists.", "evidenceIds": [ "taxjar-docs-3", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-public-api", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "TaxJar documents a full public REST API with API-key auth, versioned headers, and official client libraries in seven languages, covering rate lookups, transaction upload, and tax code parameters — clearly a documented public API for programmatic/agentic use. missing for 10: no discoverable OpenAPI/Swagger spec or llms.txt (probes returned 404s), and no independent/hands-on corroboration of API usability beyond vendor docs.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-2", "taxjar-docs-3", "taxjar-docs-4", "taxjar-docs-5", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-scoped-keys", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "TaxJar's docs only describe a single API key mechanism with no mention of scoped, least-privilege, or per-agent credential issuance; no OAuth scopes, roles, or granular permission tokens are evidenced, and OpenAPI/llms.txt probes returned 404.", "evidenceIds": [ "taxjar-docs-2", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-sdks", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "TaxJar documents official API clients across seven languages (Ruby, Python, PHP, Node, C#/.NET, Java, Go) with API key auth and versioned headers, giving AI-native developers concrete official SDKs to build against. Missing for 10: no public OpenAPI spec or llms.txt (both probes 404), and no independent/hands-on corroboration of SDK quality or completeness.", "evidenceIds": [ "taxjar-docs-3", "taxjar-docs-2", "taxjar-docs-4", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "agentic-webhooks", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of any webhook or event-subscription capability in TaxJar's API docs; only REST API key access, client libraries, and integrations are documented, with no mention of webhooks.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-2", "taxjar-docs-3" ] }, { "productId": "taxjar", "storyId": "ai-taxability-answers", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence of any AI/chatbot feature answering nexus or invoice-taxation questions with citations; evidence pack only covers API endpoints, AutoFile, nexus threshold tracking dashboards, and integrations. Probe results even show missing llms.txt/openapi discoverability artifacts, and there's no mention of an AI assistant anywhere in docs or community items.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-7", "taxjar-docs-8", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "api-interactive-docs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence of an interactive API reference with runnable examples; probes for llms.txt and OpenAPI spec both returned 404, and docs only describe static API reference content and client libraries.", "evidenceIds": [ "taxjar-probe-1", "taxjar-probe-2", "taxjar-docs-1", "taxjar-docs-3" ] }, { "productId": "taxjar", "storyId": "api-machine-spec", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Direct probes for llms.txt and common OpenAPI/Swagger spec paths all returned 404, and no documentation references a downloadable machine-readable API spec; only client libraries and reference docs are mentioned.", "evidenceIds": [ "taxjar-probe-1", "taxjar-probe-2", "taxjar-docs-1", "taxjar-docs-3" ] }, { "productId": "taxjar", "storyId": "api-sandbox", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence pack mentions a sandbox, test mode, or non-production environment for the TaxJar API; only production API usage, keys, and integrations are documented.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "api-versioning-policy", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "TaxJar's API requires an 'x-api-version' header, indicating some form of API versioning exists, but there is no documented deprecation policy, changelog, or sunset timeline evidenced anywhere in the pack. Probes for OpenAPI spec and llms.txt both failed (404), further limiting machine-readable API contract discovery for AI-native tooling. Missing for 10: documented deprecation/sunset policy, changelog of version changes, OpenAPI/machine-readable spec, and any evidence of long-term version support commitments.", "evidenceIds": [ "taxjar-docs-4", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "autofiling-remittance", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "TaxJar's AutoFile is explicitly documented to handle state tax returns and remittances with deadlines guaranteed, backed by a vendor guarantee (taxjar-docs-6), directly matching the core of the story. However, evidence doesn't clarify guarantee scope/terms (e.g., what penalties are covered, caps, exclusions), doesn't confirm coverage in every registered jurisdiction (state-only per wording), and there's no independent/hands-on corroboration of the guarantee being honored in practice. Missing for 10: detailed guarantee terms/limits, evidence of local/city jurisdiction filing beyond state level, and independent customer confirmation of guarantee payouts.", "evidenceIds": [ "taxjar-docs-6" ] }, { "productId": "taxjar", "storyId": "automation-bulk-operations", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "The API docs vaguely mention that you can \"upload transactions\" via the API, hinting at bulk transaction import, but there is no documented batch/bulk endpoint, CSV import spec, or rate-limit guidance for large-scale calls. Missing for 10: explicit bulk/batch API endpoint documentation, throughput/rate-limit specs for large batches, and any first-party or independent evidence of successful bulk operations at scale.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-5" ] }, { "productId": "taxjar", "storyId": "automation-rules-engine", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "TaxJar offers fixed automated behaviors (AutoFile filing, nexus threshold tracking) but there is no evidence of a user-configurable rules/trigger system letting an AI-native user define custom event-based automations; no webhooks, event API, or rule engine documented.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8" ] }, { "productId": "taxjar", "storyId": "automation-scheduled-jobs", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "TaxJar's evidence covers API access, AutoFile, nexus tracking, and integrations, but there is no mention of scheduling recurring jobs/workflows or any workflow-automation/cron-like feature for AI-native users. AutoFile is a fixed filing automation, not a user-schedulable job system. missing for 10: any scheduler/cron API, workflow orchestration, or recurring job configuration capability.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-1" ] }, { "productId": "taxjar", "storyId": "automation-versioned-workflows", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "TaxJar is a sales-tax calculation/filing API and SaaS product, not an automation/workflow builder; there is no concept of 'automations' to version, review, or roll back in its product surface. This axis is a category error for this product type.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "billing-stack-integrations", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "TaxJar documents pre-built integrations for major commerce/payment platforms (Amazon FBA, BigCommerce, eBay, Magento 2, NetSuite, PayPal, Shopify, Square, Stripe, WooCommerce) plus official API client libraries in seven languages, directly addressing the plug-and-play revenue-stack integration story. Missing for 10: no independent/hands-on verification that these integrations are actively maintained/up-to-date, no evidence of a public integrations marketplace or update changelog, and no OpenAPI spec available (probe returned 404s) to assess developer experience depth.", "evidenceIds": [ "taxjar-docs-9", "taxjar-docs-3", "taxjar-docs-1" ] }, { "productId": "taxjar", "storyId": "calculation-accuracy-guarantee", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The only guarantee-related evidence concerns AutoFile ensuring filing deadlines are met, not calculation accuracy or audit support for erroneous tax calculations. No documentation in the pack describes a calculation-accuracy guarantee, indemnification, or audit-support commitment tied to TaxJar's tax rate/amount outputs.", "evidenceIds": [ "taxjar-docs-6" ] }, { "productId": "taxjar", "storyId": "compliant-invoice-fields", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "TaxJar's evidence pack covers only US sales tax rates, categories, economic nexus tracking, and state tax return AutoFiling — there is no mention of VAT-specific fields, EU/international registration numbers, invoice generation, or locale-specific compliance rules for VAT jurisdictions. The axis is fair for a tax-compliance product, but no evidence shows TaxJar addresses VAT invoicing at all.", "evidenceIds": [ "taxjar-docs-5", "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-9" ] }, { "productId": "taxjar", "storyId": "exempt-customer-api", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence pack item mentions exemption certificates, exemption reasons, or marking customers/transactions exempt via the API — only generic API reference, product_tax_code, and AutoFile filing features are documented. Missing for 10: any mention of exemption_type/exemption_id parameters, exemption certificate storage/reporting, or exemption data flowing into filed returns.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "exemption-certificate-management", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "No evidence in the pack describes exemption/resale certificate collection, validation, expiry tracking, or automatic application at calculation time — the pack only covers tax rate/category APIs, AutoFile, nexus tracking, and integrations, none of which address certificate management.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "filing-calendar-visibility", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence confirms AutoFile handles filing/remittance and nexus tracking, but there is no mention of a filing calendar, per-filing status view, remitted-amount reporting, or downloadable confirmations for submitted returns. missing for 10: filing calendar UI, per-filing status tracking, remittance amount reporting, downloadable filing confirmations.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8" ] }, { "productId": "taxjar", "storyId": "global-vat-gst", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "All evidence describes US sales tax nexus tracking, AutoFile for state returns, and US-centric integrations; nothing addresses EU OSS/IOSS, UK VAT, Canada GST/HST, Australia GST, foreign VAT registration thresholds, or international filing support.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8", "taxjar-docs-9" ] }, { "productId": "taxjar", "storyId": "historical-exposure-analysis", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "Evidence shows TaxJar can track economic nexus thresholds and upload/import transactions via API, but nothing describes a backdated exposure analysis with estimated liability, penalties, and interest per jurisdiction. Missing for 10: any mention of historical liability calculation, penalty/interest estimation, or jurisdiction-level backdated exposure reporting.", "evidenceIds": [ "taxjar-docs-7", "taxjar-docs-8", "taxjar-docs-1" ] }, { "productId": "taxjar", "storyId": "liability-reporting", "verdict": "partial", "quality": 3, "confidence": "low", "rationale": "Evidence shows nexus threshold tracking and AutoFile remittance/filing automation, which implies some jurisdiction-level tax handling, but there is no documentation of a liability report showing collected-vs-owed by jurisdiction, no mention of filing-ready detail views, and no export formats designed for accountants or GL consumption. missing for 10: collected-vs-owed liability dashboard, jurisdiction-level accrued liability reporting, GL/accountant-ready export formats, evidence of real-time 'at any moment' visibility.", "evidenceIds": [ "taxjar-docs-7", "taxjar-docs-8", "taxjar-docs-6" ] }, { "productId": "taxjar", "storyId": "marketplace-facilitator-handling", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows TaxJar tracks economic nexus thresholds and integrates with Amazon FBA, but there is no documentation describing how marketplace-facilitator-collected transactions are flagged, excluded from remittance, or reported separately for nexus purposes — the exact capability the story asks about is unaddressed.", "evidenceIds": [ "taxjar-docs-7", "taxjar-docs-8", "taxjar-docs-9" ] }, { "productId": "taxjar", "storyId": "nexus-threshold-monitoring", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "TaxJar's economic nexus dashboard tracks sales/transaction counts against state thresholds and 'pinpoints states where you've met economic nexus thresholds' (taxjar-docs-7, taxjar-docs-8), showing real-time progress like '$90k of $100k' and '125 of 200' transactions. However, evidence does not confirm proactive alerting BEFORE crossing a threshold (vs. after), nor coverage of country-level (international) thresholds as the story specifies. missing for 10: explicit pre-crossing alert/notification mechanism, international/country threshold tracking, independent user confirmation of alert timing.", "evidenceIds": [ "taxjar-docs-7", "taxjar-docs-8" ] }, { "productId": "taxjar", "storyId": "omnichannel-import", "verdict": "full", "quality": 7, "confidence": "medium", "rationale": "TaxJar's API supports uploading transactions and backfilling data programmatically, and its documented integrations list covers major storefronts and processors (Amazon FBA, BigCommerce, Ebay, Magento 2, Netsuite, PayPal, Shopify, Square, Stripe, Woocommerce), enabling multi-channel data consolidation for filings. This directly supports the developer story of importing transactions from multiple sources into one nexus/filing view. Missing for 10: no independent/hands-on verification of connector sync reliability across all listed platforms, no detail on how conflicting/duplicate transactions from multiple sources are reconciled, and no OpenAPI spec confirming full API backfill capabilities (probe endpoints 404).", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-9", "taxjar-docs-7", "taxjar-docs-8", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "openness-api-parity", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Docs show a real API for tax calculation, categories, and transaction uploads, but no evidence that UI-only features like AutoFile automated filing or the economic nexus threshold dashboard are exposed via API, and probes confirm no public OpenAPI spec or llms.txt exists. missing for 10: API coverage for AutoFile/state filing, API for nexus threshold tracking/dashboard, machine-readable OpenAPI spec for full API discoverability.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-5", "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "openness-full-export", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows only API access for uploading/retrieving transaction and rate data, with no documented bulk data-export feature, open-format full account export, or data-portability tooling; probes for llms.txt and OpenAPI specs both failed (404s). Missing for 10: any documented full-account data export capability, open format (CSV/JSON) bulk export tool, or data portability/account closure export guarantee.", "evidenceIds": [ "taxjar-docs-1", "taxjar-probe-1", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "openness-open-license", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "TaxJar is a commercial SaaS tax API product, not an open-source codebase; source availability under an open license is not a fair axis for this product category.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "openness-self-host", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "TaxJar is a cloud SaaS tax-calculation/API service, not open-source software; self-hosting the core product is a category error for this type of service.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "privacy-data-residency", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence in the pack mentions data residency, regional data storage options, or any control over where TaxJar stores customer data; all documentation covers API usage, integrations, and tax filing features.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "privacy-no-training", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "TaxJar is a sales tax calculation/compliance API, not an AI model or AI-native product with data-training policies; no evidence relates to AI training data usage, so this axis is a category error for this product type.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "privacy-retention-controls", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "No evidence in the pack addresses data retention policies, deletion controls, or privacy/data lifecycle management for TaxJar's API or platform; documentation focuses solely on tax calculation, filing, and integrations.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "privacy-telemetry-optout", "verdict": "na", "quality": 0, "confidence": "high", "rationale": "TaxJar is a sales-tax compliance/API product, not an AI agent or tool with client-side telemetry settings relevant to AI-native usage; opting out of telemetry/usage tracking is not a fair axis for this product category.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "product-tax-codes", "verdict": "partial", "quality": 6, "confidence": "medium", "rationale": "TaxJar's API supports a `product_tax_code` parameter per line item, which maps catalog items to tax categories (e.g., SaaS, digital goods, services, physical goods) so each is taxed correctly per jurisdiction — this is the core mechanism for the story. However, evidence lacks a published/detailed taxonomy list of category codes, guidance on how many categories exist or how granular they are, and no independent/hands-on confirmation of correct jurisdictional tax application per category. missing for 10: full published category taxonomy/reference list, hands-on validation of correct per-jurisdiction tax outcomes by category, documentation on default categorization behavior when no code is supplied.", "evidenceIds": [ "taxjar-docs-5", "taxjar-docs-1" ] }, { "productId": "taxjar", "storyId": "rate-rule-currency", "verdict": "partial", "quality": 4, "confidence": "low", "rationale": "TaxJar's API model implies rates are served dynamically (docs-1, docs-4 x-api-version header) so calculations don't require client redeploys, and nexus/AutoFile features suggest ongoing compliance handling, but there's no explicit documentation of a rate/rule change log, update cadence, or effective-date notices for jurisdiction changes. Missing for 10: explicit changelog/documentation of when rate/rule updates take effect, confirmation of automatic jurisdiction coverage updates, and independent verification that no client-side action is ever needed.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-4", "taxjar-docs-6", "taxjar-docs-7" ] }, { "productId": "taxjar", "storyId": "realtime-tax-calculation", "verdict": "partial", "quality": 5, "confidence": "medium", "rationale": "Docs confirm a REST API with API keys, multi-language clients, and line-item/product_tax_code support for tax calculation, implying real-time rate lookups suitable for checkout integration, but no evidence explicitly addresses rooftop/address-level accuracy claims or documented latency/performance suitable for checkout timing. Missing for 10: explicit documentation of rooftop-level address accuracy vs ZIP approximation, published latency/performance benchmarks, independent/hands-on corroboration of real-time checkout use, and a working OpenAPI spec (probes returned 404s).", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-3", "taxjar-docs-4", "taxjar-docs-5", "taxjar-probe-2" ] }, { "productId": "taxjar", "storyId": "refund-adjustment-handling", "verdict": "none", "quality": 0, "confidence": "low", "rationale": "No evidence pack item addresses refunds or credit notes, how negative amounts propagate through liability/reporting, or whether they auto-reconcile into the next filing without manual fixes. Evidence only covers general API usage, tax codes, AutoFile filing, and nexus tracking — none of which speaks to refund-specific tax adjustment. missing for 10: any documentation of a refund/credit-note API or workflow, confirmation that negative line items reduce liability correctly, and evidence that refunds are reflected in AutoFile filings without manual journal entries.", "evidenceIds": [] }, { "productId": "taxjar", "storyId": "saas-digital-taxability", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence only shows a generic `product_tax_code` parameter for categorizing line items, with no documentation of specific SaaS-taxability rules by state, digital-goods nuances, bundled-offering logic, or any visible research/citations behind determinations.", "evidenceIds": [ "taxjar-docs-5" ] }, { "productId": "taxjar", "storyId": "state-registrations", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence only covers TaxJar's AutoFile service for filing state tax returns and remittances, and nexus threshold tracking — nothing about registering for sales tax permits (state/local registration filing) or per-registration pricing/status tracking. Missing for 10: any mention of permit registration service, registration pricing, or registration status tracking.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8" ] }, { "productId": "taxjar", "storyId": "transaction-audit-trail", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "The evidence pack shows TaxJar's API can return rates, categories, and accept product_tax_code parameters, but there is no documentation of a stored, exportable audit record showing the exact rate, rule, address resolution, and category used for each historical transaction's calculation. Nothing addresses retention or export of calculation provenance for audit purposes.", "evidenceIds": [ "taxjar-docs-1", "taxjar-docs-5" ] }, { "productId": "taxjar", "storyId": "vat-id-validation", "verdict": "none", "quality": 0, "confidence": "high", "rationale": "TaxJar's evidence is entirely focused on US sales tax (economic nexus, AutoFile state returns, US integrations) with no mention of VAT, EU/UK tax ID validation, official registry checks (e.g., VIES), or reverse charge mechanics for B2B cross-border sales. Nothing in the docs or community evidence addresses global VAT compliance.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8", "taxjar-docs-9" ] }, { "productId": "taxjar", "storyId": "vda-remediation", "verdict": "none", "quality": 0, "confidence": "medium", "rationale": "Evidence covers nexus tracking, AutoFile remittance, and API/integrations, but nothing addresses voluntary disclosure agreements (VDAs) or back-filing/remediation support for past-due, uncollected jurisdictions. Absence of evidence for this applicable capability means it is not shown to be delivered.", "evidenceIds": [ "taxjar-docs-6", "taxjar-docs-7", "taxjar-docs-8" ] } ]