Chargebee vs Lago

Agent access

1. ai-native userPoint an agent at llms.txt or agent-oriented docs

   weight 2 · round drawn
   Chargebee✓fullprobed9/10

   A live probe confirms llms.txt returns HTTP 200 with a proper agent-oriented summary of the API reference, and docs are also served in .md format for direct agent consumption, alongside first-party MCP server docs for agent tool access. Missing for 10: no independent/community confirmation of agents actually using llms.txt in practice.

   • [probe] “PROBE llms.txt: HTTP 200 at https://apidocs.chargebee.com/llms.txt # Chargebee API Reference > REST API reference for Chargebee Billing. Th…”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”

   proof ↗⚑ flag

   Lago✓fullprobed9/10

   Lago has a confirmed live llms.txt at the root domain (HTTP 200) plus a docs-specific llms.txt index and .md-formatted doc pages, explicitly designed for agent consumption, alongside an OpenAPI schema and MCP server for programmatic access. This directly satisfies pointing an agent at agent-oriented docs, with only independent third-party confirmation of agent usage missing for a perfect score. Missing for 10: independent/hands-on verification that an agent successfully consumes these files end-to-end.

   • [probe] “PROBE llms.txt: HTTP 200 at https://getlago.com/llms.txt # Lago > Lago is the open-source billing platform for usage-based, subscription-ba…”
   • [probe] “PROBE docs-md: HTTP 200 at https://getlago.com/docs/api-reference/intro.md > ## Documentation Index > Fetch the complete documentation index…”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”

   proof ↗⚑ flag

2. ai-native userRun the product headlessly / in CI for automation

   weight 2 · round drawn
   Chargebee~partialclaimed6/10

   Chargebee exposes a full REST API (Basic Auth, JSON) plus SDKs like Node.js, which can be scripted and called from CI pipelines without a UI, and an API Explorer confirms programmatic testing is possible outside the UI. However, there is no explicit CI/CD documentation, no official CLI, and no example pipelines or automation recipes demonstrating headless CI usage. missing for 10: CI/CD-specific docs or examples, an official CLI tool, and evidence of automated pipeline integration beyond generic API/SDK access.

   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “Node.js — /docs/sdks/nodejs - Install: npm install chargebee”
   • [claimed-docs] “The API Explorer lets you make live API calls in a browser using a guided interface. Use it to test endpoints without writing code.”

   proof ↗⚑ flag

   Lago~partialclaimed6/10

   Lago is self-hostable via Docker (docs-11), fully API/CLI-driven with a generated command for every endpoint (docs-1) and an OpenAPI 3.1 schema for scripting/client generation (docs-16), and the GitHub repo contains a script that spins up a disposable local Lago instance and runs automated pricing scenarios (gh-1), all consistent with headless/CI automation. However there is no explicit CI/headless-mode documentation, example CI pipeline config, or independent report of running Lago fully headless in a pipeline. Missing for 10: dedicated CI/headless-mode docs, a documented CI pipeline example, and independent confirmation of headless automated runs.

   • [claimed-docs] “Docker is the easiest way to get started with the self-hosted version of Lago.”
   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [github] “It starts the Lago version that matches this checkout, creates a disposable local organization, and prices three illustrative AI requests.”

   proof ↗⚑ flag

3. ai-native userPlug MCP servers into this product so it can use their tools

   weight 3 · round drawn
   Chargebee—none0/10

   Evidence only shows Chargebee exposes MCP servers for external AI clients (Cursor, Claude, ChatGPT) to consume Chargebee's own tools, i.e. Chargebee acts as the MCP server, not as a client plugging in third-party MCP servers to gain their tools. There is no documentation or example of Chargebee's AI features consuming external MCP servers. Missing for 10: any evidence of Chargebee acting as an MCP client, configuration for adding external MCP server endpoints, or first-party support for extending its AI assistant with third-party MCP tool servers.

   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [claimed-docs] “Use AI inside the Chargebee Billing UI to answer questions and complete tasks.”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago—none0/10

   The evidence shows Lago exposes its own MCP server so external AI assistants can access Lago's billing data (lago-docs-9, lago-probe-4) and has AI agents (Dunning Agent) that read Lago's own account data (lago-docs-10) — this is Lago acting as an MCP server/tool provider, not as a client that plugs in external MCP servers to gain new tools. No evidence shows Lago's agents or platform consuming/connecting to third-party MCP servers.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

4. ai-native userConnect an agent via an official MCP server

   weight 3 · round drawn
   Chargebee✓fullprobed8/10

   Chargebee documents official MCP Servers that let AI clients (Cursor, Claude, ChatGPT, Codex, etc.) connect and access selected Chargebee capabilities, plus a custom MCP server builder for tailored tool sets, confirmed via docs and a probe hit. Missing for 10: independent/hands-on third-party corroboration of the MCP server working in practice.

   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago✓fullprobed8/10

   Lago is not itself an AI agent but a billing platform, so publishing an official MCP server is a fair axis; docs explicitly confirm a dedicated MCP server giving AI assistants like Claude direct access to Lago's billing data, plus a probe confirming the page exists. Missing for 10: independent/hands-on third-party verification of the MCP server working in practice.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

5. ai-native userUse an official CLI

   weight 2 · round to Lago
   Chargebee—none0/10

   Evidence shows SDKs (Node.js), REST API, and MCP servers, but there is no mention of an official CLI tool for Chargebee. Missing for 10: any documentation or evidence of an official CLI, install instructions, or command-line usage.

   ⚑ flag

   Lago~partialclaimed4/10

   Docs mention that 'every endpoint in this reference has a generated command,' implying an official CLI exists, but there is no dedicated CLI documentation, install instructions, or AI-native framing (e.g., LLM-friendly usage, agent integration) beyond this single line. Missing for 10: explicit CLI install/setup docs, AI-native use-case description, independent confirmation of CLI usage.

   • [claimed-docs] “Every endpoint in this reference has a generated command.”

   proof ↗⚑ flag

6. ai-native userDrive the product through a documented public API

   weight 3 · round to Lago
   Chargebee✓fullprobed8/10

   Chargebee documents a full REST API with auth, JSON responses, SDKs, an API Explorer for live testing, and an llms.txt reference confirming API v2 coverage — this is a clearly documented, drivable public API. Missing for 10: publicly hosted OpenAPI/swagger spec (probe found 404s) and independent third-party corroboration of API robustness.

   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “Node.js — /docs/sdks/nodejs - Install: npm install chargebee”
   • [claimed-docs] “The API Explorer lets you make live API calls in a browser using a guided interface. Use it to test endpoints without writing code.”
   • [probe] “PROBE llms.txt: HTTP 200 at https://apidocs.chargebee.com/llms.txt # Chargebee API Reference > REST API reference for Chargebee Billing. Th…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://apidocs.chargebee.com/openapi.json, https://apidocs.chargebee.com/swagger.json, https://apid…”

   proof ↗⚑ flag

   Lago✓fullprobed9/10

   Lago publishes a full public API reference with generated commands per endpoint, an OpenAPI 3.1 schema (verified live at /openapi.json), explicit versioning/stability guarantees, and docs/llms.txt aimed at AI-native consumption; it even documents an MCP server exposing billing data to AI agents. This is strong first-party and probe-verified evidence of a documented, machine-consumable public API. Missing for 10: independent third-party developer testimonials specifically about API usage/integration experience (community evidence is mostly about billing product value, not API DX).

   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “Clients should tolerate additive changes within v1, including new response fields, enum values, endpoints, optional request fields, and webh…”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “PROBE llms.txt: HTTP 200 at https://getlago.com/llms.txt # Lago > Lago is the open-source billing platform for usage-based, subscription-ba…”
   • [probe] “PROBE docs-md: HTTP 200 at https://getlago.com/docs/api-reference/intro.md > ## Documentation Index > Fetch the complete documentation index…”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”
   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

7. ai-native userIssue scoped/least-privilege API credentials for an agent

   weight 2 · round to Chargebee
   Chargebee~partialprobed4/10

   Chargebee's API uses HTTP Basic Auth with API keys and offers custom MCP servers that can be scoped to a subset of the tool catalog (e.g., invoice-only), suggesting some ability to limit an agent's surface area, but there is no documented support for scoped/least-privilege API key roles, granular permission scopes, or agent-specific credential issuance in the core REST API. missing for 10: explicit API key roles/permission scopes documentation, least-privilege credential issuance workflow for agents, independent verification that custom MCP server scoping enforces true least-privilege access control.

   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago—none0/10

   Lago's docs describe an MCP server and API access for AI agents but provide no evidence of scoped or least-privilege API key/credential issuance (e.g., role-based keys, permission scopes) for agents; existing docs only mention general API keys and OpenAPI schema access without granular scoping controls.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

8. ai-native userBuild against official SDKs

   weight 2 · round to Chargebee
   Chargebee~partialprobed6/10

   Chargebee documents official SDKs (e.g., Node.js via npm) and a REST API with docs, satisfying the core 'build against official SDKs' ask, but evidence only lists one SDK explicitly and lacks broader SDK language coverage, versioning, or independent developer corroboration; OpenAPI spec discovery also failed in probes. Missing for 10: multi-language SDK list/documentation depth, independent hands-on validation, discoverable OpenAPI/schema artifacts.

   • [claimed-docs] “Node.js — /docs/sdks/nodejs - Install: npm install chargebee”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://apidocs.chargebee.com/openapi.json, https://apidocs.chargebee.com/swagger.json, https://apid…”

   proof ↗⚑ flag

   Lago~partialprobed5/10

   Lago publishes an OpenAPI 3.1 spec and API reference with per-endpoint generated commands, explicitly positioned for generating clients or exposing actions to AI agents, but the evidence never names concrete official first-party SDKs (e.g., language libraries) beyond spec-generated clients. missing for 10: named official SDK packages/libraries, independent developer confirmation of SDK usage, versioning/maintenance details for SDKs.

   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”
   • [claimed-docs] “Clients should tolerate additive changes within v1, including new response fields, enum values, endpoints, optional request fields, and webh…”

   proof ↗⚑ flag

9. ai-native userSubscribe to events via webhooks

   weight 2 · round drawn
   Chargebee✓fullcommunity7/10

   Chargebee documents event-driven webhooks (events trigger configured webhooks) and community evidence confirms multiple webhook URLs can be configured for different environments, showing real-world subscription-to-events usage. missing for 10: no detailed webhook payload/signature-verification docs, no explicit AI-agent-oriented subscription workflow beyond generic webhook config, and no independent hands-on verification beyond the one HN mention.

   • [claimed-docs] “If webhooks are configured in Chargebee, events trigger those webhooks.”
   • [community] “With Chargebee I can add multiple webhook URLs, so all events get sent to both dev and staging and each environment only cares about its own…”

   proof ↗⚑ flag

   Lago✓fullclaimed7/10

   Lago documents a webhook system with HMAC signing options (lago-docs-7) and explicit API versioning guarantees for webhook event types (lago-docs-8), meaning any consumer—including an AI agent—can subscribe to billing events via webhooks. missing for 10: a full catalog/list of supported webhook event types, and independent/hands-on confirmation of webhook reliability in production use.

   • [claimed-docs] “The second is `HMAC`, which features a shorter payload header and no size restrictions.”
   • [claimed-docs] “Clients should tolerate additive changes within v1, including new response fields, enum values, endpoints, optional request fields, and webh…”

   proof ↗⚑ flag

Agentic features

1. ai-native userGet AI-generated insights and suggestions from my data inside the product

   weight 2 · round to Chargebee
   Chargebee~partialclaimed6/10

   Chargebee docs describe an in-product AI assistant ('Use AI inside the Chargebee Billing UI to answer questions and complete tasks') and AI that can draft plans/pricing from documents, which are in-product AI-driven capabilities, but there's no evidence of proactive AI-generated insights or suggestions derived from the customer's own billing/subscription data (e.g., churn risk, revenue anomalies) — the described features skew toward Q&A and configuration drafting rather than data-driven insight generation. Missing for 10: evidence of analytics-style insights/suggestions surfaced from account data, and independent/hands-on validation of the AI assistant's usefulness.

   • [claimed-docs] “Use AI inside the Chargebee Billing UI to answer questions and complete tasks.”
   • [claimed-docs] “Let Chargebee AI draft product families, plans, addons, charges, prices, features, and entitlements from a pricing page, document, or descri…”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”

   proof ↗⚑ flag

   Lago~partialprobed5/10

   Lago documents specific AI-agent features—an MCP server that gives assistants like Claude direct access to billing data, and a 'Dunning Agent' that reads 12 months of invoices/payments/disputes to inform actions—showing some in-product AI analysis of data. However, this is narrow (payment collection) rather than a general AI-insights/suggestions layer across revenue analytics, plans, or usage data. missing for 10: evidence of AI-generated insights/suggestions surfaced in the main product UI (e.g., revenue analytics, plan optimization), independent/hands-on confirmation the Dunning Agent's outputs are used, and broader scope beyond dunning/collections.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

2. ai-native userSet up automations that run autonomously in the background

   weight 2 · round to Lago
   Chargebee~partialprobed4/10

   Chargebee offers webhooks, dunning/smart retry, and MCP servers that let AI clients access billing capabilities, which could underpin background automations, but there's no direct evidence of a scheduler or agent-orchestration feature that runs autonomous, unattended workflows over time. missing for 10: explicit autonomous/background automation or workflow engine documentation, evidence of scheduled/triggered agentic tasks, independent confirmation of unattended operation.

   • [claimed-docs] “If webhooks are configured in Chargebee, events trigger those webhooks.”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [claimed-docs] “With Chargebee's Dunning feature, you can send email reminders to your customers asking them to update payment details and schedule payment …”
   • [claimed-docs] “Chargebee's Smart retry logic retries payments at dynamic intervals based on the type of gateway transaction errors, and it acts on the hard…”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago~partialprobed6/10

   Lago documents a Dunning Agent that autonomously reads account/invoice/payment history and presumably acts on overdue accounts, plus an MCP server exposing billing data to AI agents like Claude, showing some autonomous background automation capability. However, evidence doesn't show a general-purpose automation/agent-builder framework, scheduling, or independent confirmation of the Dunning Agent running unattended in production. missing for 10: broader autonomous automation framework beyond dunning agent, independent/hands-on verification of autonomous background operation, details on triggering/scheduling/monitoring these agents.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

3. ai-native userDelegate tasks to a built-in AI assistant inside the product

   weight 3 · round to Chargebee
   Chargebee~partialclaimed6/10

   Chargebee documents a built-in AI assistant inside the Billing UI that can 'answer questions and complete tasks' and draft pricing/product configurations from a description, showing real in-product task delegation (chargebee-docs-16, chargebee-docs-14). However, the primary AI-native evidence emphasizes MCP servers for external AI clients (Cursor, Claude, ChatGPT) rather than a deeply documented native assistant experience, and there is no independent/hands-on corroboration of the in-app assistant's task scope or reliability. missing for 10: independent verification of the built-in assistant's capabilities, detail on task types beyond drafting/answering, and community evidence confirming real-world use.

   • [claimed-docs] “Use AI inside the Chargebee Billing UI to answer questions and complete tasks.”
   • [claimed-docs] “Let Chargebee AI draft product families, plans, addons, charges, prices, features, and entitlements from a pricing page, document, or descri…”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”

   proof ↗⚑ flag

   Lago~partialprobed5/10

   Lago ships an 'AI agents' documentation section describing a built-in Dunning Agent that autonomously reads invoices, disputes, and payment provider status to act on overdue accounts, and an MCP server that gives external AI assistants (e.g., Claude) direct access to billing data. This shows some built-in agentic delegation for a specific task (dunning/collections), but there's no evidence of a general-purpose in-product AI assistant users can converse with or delegate arbitrary tasks to. missing for 10: evidence of a general conversational assistant UI, breadth of tasks beyond dunning, independent/hands-on confirmation of the Dunning Agent actually operating in production.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

4. ai-native userOperate the product with natural-language commands

   weight 2 · round drawn
   Chargebee~partialprobed6/10

   Chargebee documents an in-product AI assistant that lets users 'answer questions and complete tasks' via natural language inside the Billing UI, plus AI-driven drafting of pricing/plans from a description, and MCP servers that let external AI clients (Cursor, Claude, ChatGPT) operate Chargebee capabilities conversationally. This is first-party doc evidence only, with no independent/hands-on corroboration of how robust or complete the natural-language command coverage actually is. Missing for 10: independent/hands-on validation of the AI assistant's command coverage, and detail on which actions can be fully executed vs. only queried via natural language.

   • [claimed-docs] “Use AI inside the Chargebee Billing UI to answer questions and complete tasks.”
   • [claimed-docs] “Let Chargebee AI draft product families, plans, addons, charges, prices, features, and entitlements from a pricing page, document, or descri…”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago~partialprobed6/10

   Lago documents an official MCP server that gives AI assistants like Claude direct access to billing data, plus specialized AI agents (e.g., Dunning Agent) that read account data and OpenAPI schema exposure for AI agents to invoke billing actions — enabling natural-language operation via connected AI assistants. However, this is mediated through external assistants/MCP rather than a built-in NL command interface, and there's no independent/hands-on evidence of actual NL command execution or the breadth of supported commands. Missing for 10: hands-on/independent verification of NL command execution, a native in-product chat/command interface, and documentation of the full range of commands executable via natural language.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

Api quality

1. ai-native userExplore an interactive API reference with runnable examples

   weight 2 · round to Chargebee
   Chargebee~partialprobed6/10

   Chargebee documents an API Explorer that lets users make live API calls in a browser with a guided interface to test endpoints without writing code, which matches the interactive/runnable-example story. However, no OpenAPI spec was discoverable via probe (all candidate paths 404), and there's no independent/hands-on corroboration of the explorer's depth or coverage. Missing for 10: discoverable OpenAPI/Swagger spec, independent/hands-on validation of the API Explorer's runnable examples across endpoints.

   • [claimed-docs] “The API Explorer lets you make live API calls in a browser using a guided interface. Use it to test endpoints without writing code.”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://apidocs.chargebee.com/openapi.json, https://apidocs.chargebee.com/swagger.json, https://apid…”

   proof ↗⚑ flag

   Lago~partialprobed5/10

   Docs confirm a generated command for every endpoint and an OpenAPI 3.1 schema (lago-docs-1, lago-docs-16, lago-probe-3), suggesting a structured, code-snippet-rich API reference, but there is no evidence of an in-browser 'try it out' or runnable execution feature. Missing for 10: explicit interactive 'run this request' UI, hands-on confirmation of executability, and independent corroboration of the reference's interactivity.

   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”
   • [probe] “PROBE docs-md: HTTP 200 at https://getlago.com/docs/api-reference/intro.md > ## Documentation Index > Fetch the complete documentation index…”

   proof ↗⚑ flag

2. ai-native userDownload a machine-readable API spec (OpenAPI or equivalent)

   weight 2 · round to Lago
   Chargebee—none0/10

   Direct probes for standard OpenAPI/Swagger spec locations all returned 404, and no documentation page claims a downloadable OpenAPI/machine-readable spec exists; only an llms.txt index and human-readable markdown docs were found.

   • [probe] “PROBE openapi: all candidate paths 404 (https://apidocs.chargebee.com/openapi.json, https://apidocs.chargebee.com/swagger.json, https://apid…”
   • [probe] “PROBE llms.txt: HTTP 200 at https://apidocs.chargebee.com/llms.txt # Chargebee API Reference > REST API reference for Chargebee Billing. Th…”
   • [probe] “PROBE docs-md: HTTP 404 at https://apidocs.chargebee.com/docs/api/getting-started.md.md”

   proof ↗⚑ flag

   Lago✓fullprobed9/10

   Lago publishes an OpenAPI 3.1 schema (openapi.json returns 200 with an 'openapi' key) and docs explicitly state it can be used to generate clients, inspect typed operations, or expose actions to an AI agent. missing for 10: no independent third-party corroboration of the spec's completeness or usage beyond vendor docs/probe.

   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”

   proof ↗⚑ flag

3. ai-native userTest against a sandbox environment without touching production data

   weight 1 · round drawn
   Chargebee~partialcommunity4/10

   Evidence shows Chargebee supports separate dev/staging environments (webhooks routed independently per environment) and an API Explorer for testing endpoints without code, implying some sandbox-like testing capability, but there is no explicit documentation of a dedicated sandbox/test site that guarantees production data isolation for AI-driven testing. Missing for 10: explicit sandbox/test-mode documentation, confirmation that API keys or MCP tools can target a non-production environment, and first-party guidance on avoiding production data during AI testing.

   • [claimed-docs] “The API Explorer lets you make live API calls in a browser using a guided interface. Use it to test endpoints without writing code.”
   • [community] “With Chargebee I can add multiple webhook URLs, so all events get sent to both dev and staging and each environment only cares about its own…”

   proof ↗⚑ flag

   Lago~partialclaimed4/10

   Lago's self-hosted Docker setup and the GitHub quick-start script that spins up a 'disposable local organization' for testing suggest a way to try Lago without touching production data, but there is no documented dedicated sandbox/test-mode with separate test API keys as many billing platforms provide. Missing for 10: explicit sandbox/test-mode documentation, test vs. live API key separation, and hands-on confirmation that test data is fully isolated from production billing.

   • [github] “It starts the Lago version that matches this checkout, creates a disposable local organization, and prices three illustrative AI requests.”
   • [claimed-docs] “Docker is the easiest way to get started with the self-hosted version of Lago.”

   proof ↗⚑ flag

4. ai-native userRely on versioned APIs with a documented deprecation policy

   weight 2 · round to Lago
   Chargebee~partialprobed3/10

   Evidence shows Chargebee's API is versioned (API v2, Product Catalog) per the llms.txt probe, but there is no documentation of a deprecation policy, sunset timeline, or migration guidance for older API versions anywhere in the pack. Missing for 10: explicit deprecation/sunset policy docs, versioning changelog, migration guide for deprecated endpoints, independent confirmation of long-term API stability.

   • [probe] “PROBE llms.txt: HTTP 200 at https://apidocs.chargebee.com/llms.txt # Chargebee API Reference > REST API reference for Chargebee Billing. Th…”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”

   proof ↗⚑ flag

   Lago✓fullprobed8/10

   Lago documents an explicit API versioning policy (v1) specifying that clients should tolerate additive changes like new fields, enums, endpoints, and webhook event types, indicating a stable versioned API contract with deprecation-safe evolution rules, and backs this with a published OpenAPI 3.1 schema. Missing for 10: an explicit deprecation/sunset timeline policy (e.g., how long old versions are supported) and independent/community confirmation of version stability in practice.

   • [claimed-docs] “Clients should tolerate additive changes within v1, including new response fields, enum values, endpoints, optional request fields, and webh…”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”

   proof ↗⚑ flag

1. ai-native userPerform bulk operations across many items at once

   weight 2 · round to Chargebee
   Chargebee~partialprobed5/10

   Chargebee's REST API, SDKs, and Export resource allow programmatic access to many records at once (e.g. bulk export jobs), and its MCP servers let AI clients invoke these API operations, but the evidence never documents a dedicated bulk/batch-update endpoint for mutating many items in a single call — automation would rely on looping single-item API calls. missing for 10: explicit bulk-create/bulk-update API endpoints, documented batch size limits, and independent confirmation of bulk operations working via MCP/AI clients.

   • [claimed-docs] “Export resource represents an export job and contains the status of the job and the download URL, if the job is successfully completed.”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “Node.js — /docs/sdks/nodejs - Install: npm install chargebee”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago—none0/10

   No evidence describes bulk/batch endpoints or batch operations (e.g., bulk invoice creation, bulk customer updates) in Lago's API or UI; documentation only mentions per-endpoint commands and general API access. Missing for 10: any mention of batch/bulk API endpoints, batch job status, or UI batch-select actions.

   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”

   proof ↗⚑ flag

2. ai-native userDefine rules that trigger actions automatically on events

   weight 3 · round drawn
   Chargebee~partialcommunity5/10

   Chargebee supports event-driven automation via webhooks (events trigger webhooks, multiple URLs supported) and has built-in automated actions like Dunning email reminders and Smart Retry logic that act on transaction event types, but there's no evidence of a general-purpose, user-configurable rules/workflow engine where an AI-native user can define arbitrary 'if event X then action Y' rules beyond these pre-built billing automations. Missing for 10: a documented rules/workflow builder or automation engine, evidence of custom trigger-action definitions beyond webhooks/dunning/retries, and independent confirmation of flexibility.

   • [claimed-docs] “If webhooks are configured in Chargebee, events trigger those webhooks.”
   • [community] “With Chargebee I can add multiple webhook URLs, so all events get sent to both dev and staging and each environment only cares about its own…”
   • [claimed-docs] “With Chargebee's Dunning feature, you can send email reminders to your customers asking them to update payment details and schedule payment …”
   • [claimed-docs] “Chargebee's Smart retry logic retries payments at dynamic intervals based on the type of gateway transaction errors, and it acts on the hard…”

   proof ↗⚑ flag

   Lago~partialclaimed5/10

   Lago documents webhooks that fire on billing events (lago-docs-7) and a dedicated 'Dunning Agent' that reads account state (invoices, payments, disputes) and presumably acts on overdue-payment events (lago-docs-10), which together resemble event-triggered automation. However, there is no documented general-purpose rule/condition engine letting users define arbitrary 'if event X then action Y' logic beyond these specific built-in agents/webhooks. Missing for 10: a configurable rules/conditions UI or API for arbitrary triggers, and evidence of multiple event types beyond dunning/webhooks.

   • [claimed-docs] “The second is `HMAC`, which features a shorter payload header and no size restrictions.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”

   proof ↗⚑ flag

3. ai-native userSchedule recurring jobs or workflows

   weight 2 · round to Lago
   Chargebee—none0/10

   Chargebee's evidence covers recurring billing cycles, dunning retry schedules, and async export jobs, but these are internal billing mechanics, not a user- or AI-schedulable job/workflow automation feature (e.g., no API endpoint or MCP tool to create/schedule arbitrary recurring workflows). The MCP server evidence describes connecting AI clients to Chargebee's data/tools, not job scheduling.

   • [claimed-docs] “With Chargebee's Dunning feature, you can send email reminders to your customers asking them to update payment details and schedule payment …”
   • [claimed-docs] “Chargebee's Smart retry logic retries payments at dynamic intervals based on the type of gateway transaction errors, and it acts on the hard…”
   • [claimed-docs] “Multi-frequency billing: Subscriptions are billed according to the billing period of each recurring item price within the subscription, whet…”
   • [claimed-docs] “Export resource represents an export job and contains the status of the job and the download URL, if the job is successfully completed.”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”

   proof ↗⚑ flag

   Lago~partialclaimed3/10

   Lago's core billing engine inherently runs recurring cycles (monthly/yearly subscription billing) and the Dunning Agent appears to run periodic checks over invoices/payments, which is a recurring automated workflow, but there is no documented feature letting an AI-native user define or schedule arbitrary recurring jobs/workflows via API or agent. missing for 10: explicit job/workflow scheduler, cron-like API, user-configurable recurrence for AI agent tasks beyond built-in billing cycles.

   • [claimed-docs] “Yearly: subscription fees are billed on a yearly basis and charges can be billed monthly or annually.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”

   proof ↗⚑ flag

Agent operations

1. ai-native userAn agent can operate my billing — create a subscription, apply a credit, change a plan — through a programmatic surface with scoped permissions

   weight 3 · round to Lago
   Chargebee~partialprobed6/10

   Chargebee has a full REST API (with Basic Auth, SDKs) that supports subscription creation, plan changes, and credits, plus an official MCP server that exposes 'selected capabilities' to AI clients like Cursor/Claude/ChatGPT, and custom MCP servers that scope which tools are exposed. This directly supports agent-driven billing operations with some permission scoping, but evidence doesn't confirm granular scoped-permission models (e.g., per-action API keys) for agents beyond curated tool subsets, nor independent/hands-on validation of agent-driven write operations. missing for 10: documented fine-grained scoped API-key/OAuth permission model for agent use, independent/hands-on evidence of an agent successfully executing subscription/credit/plan-change operations via MCP or API.

   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “An item price is compatible with a plan-item price if their currencies are the same.”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [claimed-docs] “Use AI inside the Chargebee Billing UI to answer questions and complete tasks.”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago~partialprobed7/10

   Lago exposes a full REST API (every endpoint has a generated command) plus an OpenAPI 3.1 schema explicitly positioned for exposing billing actions to AI agents, and ships a documented MCP server giving AI assistants direct access to billing data, with a working example (Dunning Agent) reading account/invoice data. Subscription assignment and plan upgrade/downgrade are documented API-driven flows, supporting the 'create a subscription, change a plan' part of the story. Missing for 10: no explicit documentation of scoped/permissioned API keys or credit-application endpoints for agents, and no independent hands-on report of an agent actually performing these billing operations end-to-end.

   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “Customers have the flexibility to easily upgrade or downgrade their plan at any given moment.”
   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”

   proof ↗⚑ flag

Agent queries

1. ai-native userAn agent can answer questions about my billing state — MRR, delinquent accounts, upcoming renewals, a customer's invoices — by querying the platform directly

   weight 2 · round to Lago
   Chargebee✓fullprobed7/10

   Chargebee offers a REST API for querying subscriptions, invoices, and account data (chargebee-docs-1) plus official MCP servers letting AI clients like Cursor, Claude, ChatGPT and Codex 'securely access selected capabilities' including subscription/invoice lookups (chargebee-docs-7, chargebee-docs-8, chargebee-probe-4), directly enabling agent-driven billing queries. Missing for 10: explicit documented support for MRR/delinquent-account aggregate queries via MCP or API, and independent/hands-on confirmation that the MCP server reliably answers such billing-state questions.

   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “Connect AI clients such as Cursor, Claude, ChatGPT, Codex, and other MCP-compatible tools to Chargebee MCP Servers to securely access select…”
   • [claimed-docs] “Custom MCP servers let you combine tools from Chargebee's full tool catalog into a server tailored to your use case—for example, an invoice …”
   • [claimed-docs] “Use AI inside the Chargebee Billing UI to answer questions and complete tasks.”
   • [probe] “official MCP server documented at https://www.chargebee.com/docs/billing/2.0/ai-in-chargebee/chargebee-mcp”

   proof ↗⚑ flag

   Lago✓fullprobed8/10

   Lago ships an official MCP server explicitly described as giving AI assistants 'direct access to Lago's billing data' (lago-docs-9, lago-probe-4), plus a documented OpenAPI schema explicitly positioned for 'exposing billing actions to an AI agent' (lago-docs-16), and a Dunning Agent example that reads a customer's invoices, payments and disputes (lago-docs-10) — directly matching the story of an agent querying billing state like delinquency and invoices. Missing for 10: explicit worked examples/queries for MRR aggregation and upcoming-renewal lookups via the MCP server, and independent (non-vendor) confirmation that these agent queries work in practice.

   • [claimed-docs] “provides AI assistants (like Claude) with direct access to Lago’s billing data.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “official MCP server documented at https://getlago.com/docs/guide/ai-agents/mcp-server”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”

   proof ↗⚑ flag

Billing cycles

1. ops userControl billing anchors and cycles — calendar vs anniversary billing, alignment across a customer's subscriptions, custom invoice dates

   weight 2 · round drawn
   Chargebee—none0/10

   The evidence pack contains no documentation on billing anchor dates, calendar-vs-anniversary billing choice, cross-subscription date alignment, or custom invoice date control — only generic multi-frequency billing (chargebee-docs-15) which doesn't address anchor/alignment mechanics. Missing for 10: docs on billing anchor/alignment settings, calendar vs anniversary billing options, custom invoice date configuration.

   • [claimed-docs] “Multi-frequency billing: Subscriptions are billed according to the billing period of each recurring item price within the subscription, whet…”

   proof ↗⚑ flag

   Lago—none0/10

   The evidence pack shows general plan/subscription docs (yearly billing, upgrades/downgrades) but never addresses calendar vs anniversary billing anchors, cross-subscription date alignment, or custom invoice date overrides — the specific billing-math controls this story asks about.

   • [claimed-docs] “Yearly: subscription fees are billed on a yearly basis and charges can be billed monthly or annually.”
   • [claimed-docs] “In the “Overview” tab, click “Add a plan” on the right”
   • [claimed-docs] “Customers have the flexibility to easily upgrade or downgrade their plan at any given moment.”

   proof ↗⚑ flag

Proration

1. finance leadMid-cycle upgrades and downgrades are prorated with documented, auditable math — credits, partial periods, and invoice line items that reconcile to the penny

   weight 3 · round to Lago
   Chargebee—none0/10

   The evidence pack contains only generic API/billing feature docs (usage events, item price compatibility, dunning, RevRec) but nothing about proration mechanics, credit calculations, partial-period invoice line items, or auditable math for upgrades/downgrades. Missing for 10: proration policy docs, credit note reconciliation examples, partial-period invoice line item breakdowns, and any audit-trail evidence.

   ⚑ flag

   Lago~partialcommunity3/10

   Lago documents that customers can upgrade/downgrade plans at any time (lago-docs-4) and has dedicated subscription/plan docs, implying some proration handling, but no evidence shows documented proration formulas, credit line-item detail, or penny-level invoice reconciliation. A community question about upgrade/downgrade tracking (lago-comm-3) goes unanswered in the evidence pack. Missing for 10: explicit proration algorithm documentation, sample invoice line items showing credits/partial periods, and any audit or reconciliation guarantee.

   • [claimed-docs] “Customers have the flexibility to easily upgrade or downgrade their plan at any given moment.”
   • [claimed-docs] “In the “Overview” tab, click “Add a plan” on the right”
   • [community] “User describes building an internal billing tool after bad experiences with Recurly and ChargeBee for B2B/B2E usage-based billing, discounts…”

   proof ↗⚑ flag

Card recovery

1. ops userExpiring or replaced cards update themselves — network account updater, backup payment methods, and hosted update-payment pages

   weight 2 · round to Chargebee
   Chargebee~partialclaimed5/10

   Chargebee documents dunning email reminders and smart retry logic for recovering failed payments, which supports part of the dunning-recovery workflow, but the evidence pack contains no explicit mention of network account updater, backup payment methods, or hosted update-payment pages. missing for 10: explicit documentation of network account updater (card auto-refresh via card networks), backup/secondary payment method fallback, and hosted payment page update flows for customers.

   • [claimed-docs] “With Chargebee's Dunning feature, you can send email reminders to your customers asking them to update payment details and schedule payment …”
   • [claimed-docs] “Chargebee's Smart retry logic retries payments at dynamic intervals based on the type of gateway transaction errors, and it acts on the hard…”

   proof ↗⚑ flag

   Lago—none0/10

   No evidence in the pack mentions network account updater, backup payment methods, or hosted update-payment pages; Lago's dunning docs focus on invoice/dispute tracking and an AI dunning agent, not card-updating mechanics.

   ⚑ flag

Dunning comms

1. ops userConfigure automated dunning sequences — branded emails, in-app banners, grace periods, and final-state rules like cancel vs unpaid

   weight 2 · round to Chargebee
   Chargebee~partialclaimed5/10

   Docs confirm Chargebee's Dunning feature sends email reminders and schedules smart payment retries, covering core automated dunning emails and retry logic, but there is no evidence of in-app banners, configurable grace periods, or explicit cancel-vs-unpaid final-state rule configuration. missing for 10: in-app banner support, grace period configuration, final-state (cancel vs unpaid) rule documentation, branding/customization details for dunning emails.

   • [claimed-docs] “With Chargebee's Dunning feature, you can send email reminders to your customers asking them to update payment details and schedule payment …”
   • [claimed-docs] “Chargebee's Smart retry logic retries payments at dynamic intervals based on the type of gateway transaction errors, and it acts on the hard…”

   proof ↗⚑ flag

   Lago~partialclaimed4/10

   Lago documents dunning-adjacent capabilities: a 'cash-collection' feature to spot overdue/unpaid invoices via filters (lago-docs-13) and a 'Dunning Agent' that reads invoices, payments, disputes, and provider status (lago-docs-10). However, there is no evidence of configurable branded emails, in-app banners, grace periods, or explicit cancel-vs-unpaid final-state rules. Missing for 10: branded email templates, in-app banner UI, grace period configuration, explicit final-state (cancel/unpaid) rule settings, and independent confirmation of the dunning workflow in practice.

   • [claimed-docs] “Spot overdue, partially paid or unpaid invoices with filters to never lose track of what you’re owed.”
   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”

   proof ↗⚑ flag

Smart retries

1. finance leadFailed renewal payments are retried automatically on an ML- or schedule-optimized cadence that measurably recovers involuntary churn

   weight 3 · round to Chargebee
   Chargebee~partialclaimed6/10

   Chargebee's docs confirm scheduled dunning retries and "Smart Retry" logic that varies intervals dynamically based on gateway decline type, directly supporting automated, optimized retry cadences aimed at recovering failed renewal payments. However, missing for 10: no quantified recovery-rate data, no explicit claim of ML-based optimization (only 'dynamic intervals based on error type'), and no independent/customer evidence measuring actual involuntary-churn recovery impact.

   • [claimed-docs] “With Chargebee's Dunning feature, you can send email reminders to your customers asking them to update payment details and schedule payment …”
   • [claimed-docs] “Chargebee's Smart retry logic retries payments at dynamic intervals based on the type of gateway transaction errors, and it acts on the hard…”

   proof ↗⚑ flag

   Lago~partialclaimed4/10

   Lago documents cash-collection tooling to flag overdue/unpaid invoices and an AI 'Dunning Agent' that reads account history (invoices, payments, disputes, provider status), implying some automated dunning capability, but there is no documented ML- or schedule-optimized retry cadence, no explicit retry logic, and no metrics on churn recovery. Missing for 10: explicit description of automated retry scheduling/cadence optimization, evidence of ML-driven timing, and quantified recovery/churn-reduction outcomes.

   • [claimed-docs] “The Dunning Agent reads the account first: twelve months of invoices and payments, the disputes, the provider status.”
   • [claimed-docs] “Spot overdue, partially paid or unpaid invoices with filters to never lose track of what you’re owed.”

   proof ↗⚑ flag

Entitlement checks

1. developerFeatures and limits are modeled as entitlements attached to plans, and my application can check a customer's entitlements at runtime via API or SDK

   weight 3 · round to Chargebee
   Chargebee~partialclaimed4/10

   Chargebee documents an entitlements/Features model attached to plans, including a UI to manage customer privileges and AI-assisted entitlement creation, and separately documents a general REST API and Node.js SDK. However, no evidence explicitly shows an API/SDK endpoint or method for checking a customer's entitlements at runtime. missing for 10: explicit entitlements-check API endpoint or SDK method, code sample showing runtime lookup, independent/hands-on confirmation of this specific capability.

   • [claimed-docs] “Chargebee provides you an efficient UI to allocate and manage the customer privileges using Features.”
   • [claimed-docs] “Let Chargebee AI draft product families, plans, addons, charges, prices, features, and entitlements from a pricing page, document, or descri…”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “Node.js — /docs/sdks/nodejs - Install: npm install chargebee”

   proof ↗⚑ flag

   Lago~partialclaimed3/10

   Lago has a dedicated entitlements platform page combining billing and feature flags, confirming the concept exists, but the evidence pack contains no detail on how entitlements are attached to plans nor concrete API/SDK examples for runtime entitlement checks. missing for 10: API reference for entitlement endpoints, SDK code samples for checking a customer's entitlements at runtime, plan-to-entitlement modeling details.

   • [claimed-docs] “Combine billing and feature flags to avoid bugs and dependencies, reduce complexity and lower and engineering overhead.”

   proof ↗⚑ flag

Plan versioning

1. ops userChange pricing without breaking existing subscribers — versioned plans, grandfathering, and bulk migration tools for moving cohorts to new prices

   weight 2 · round to Lago
   Chargebee—none0/10

   Chargebee's docs describe item prices, subscriptions, and entitlements, but none of the evidence mentions versioned plans, grandfathering existing subscribers, or bulk cohort migration tools for pricing changes. Missing for 10: explicit versioned-plan mechanism, grandfathering documentation, and a bulk subscription migration tool.

   ⚑ flag

   Lago~partialclaimed3/10

   Docs show basic plan/subscription management (assign plan, upgrade/downgrade) but no explicit mention of plan versioning, grandfathering existing subscribers, or bulk cohort migration tools. missing for 10: explicit versioned-plan mechanism, grandfathering documentation, bulk/cohort migration tooling or API endpoints.

   • [claimed-docs] “In the “Overview” tab, click “Add a plan” on the right”
   • [claimed-docs] “Customers have the flexibility to easily upgrade or downgrade their plan at any given moment.”
   • [claimed-docs] “Yearly: subscription fees are billed on a yearly basis and charges can be billed monthly or annually.”

   proof ↗⚑ flag

Invoices

1. ops userThe platform generates compliant invoices, credit notes, and receipts — sequential numbering, custom fields, memos — delivered to customers automatically

   weight 2 · round to Lago
   Chargebee—none0/10

   The evidence pack contains no documentation of invoice/credit note/receipt generation, sequential numbering, custom fields, or memos — only tangential billing topics like dunning, revenue recognition, and API basics. missing for 10: invoice generation docs, credit note/receipt automation, sequential numbering compliance, custom fields/memos support.

   ⚑ flag

   Lago~partialclaimed4/10

   Lago's docs confirm core invoicing functionality (downloadable invoices, customizable invoice templates) and cash-collection tracking of overdue invoices, showing invoicing is a first-class feature, but the evidence pack contains no mention of credit notes, receipts, sequential numbering, custom fields, or memos, nor confirmation of fully automated delivery to customers. missing for 10: credit note support, receipt generation, sequential invoice numbering, custom fields/memos on invoices, automatic delivery mechanism confirmation.

   • [claimed-docs] “You can download invoices or use the invoice object to create your own invoice template.”
   • [claimed-docs] “Spot overdue, partially paid or unpaid invoices with filters to never lose track of what you’re owed.”

   proof ↗⚑ flag

Multi currency

1. finance leadBill customers in their local currency with per-currency price points and consolidated reporting in my home currency

   weight 2 · round to Chargebee
   Chargebee~partialclaimed3/10

   Chargebee docs mention item prices tied to specific currencies (implying per-currency price points exist), but there is no evidence describing consolidated reporting back into a home currency for finance leads. Missing for 10: explicit multi-currency pricing catalog documentation, currency conversion/consolidation reporting features, and any finance-lead-facing reporting tool showing home-currency rollups.

   • [claimed-docs] “An item price is compatible with a plan-item price if their currencies are the same.”

   proof ↗⚑ flag

   Lago—none0/10

   No evidence in the pack addresses multi-currency billing, per-currency price points, or consolidated home-currency reporting; docs cover plans, invoicing, dunning, and revenue analytics but never mention currency handling or FX consolidation. Missing for 10: any documentation of multi-currency price points, currency conversion/consolidation logic, and reporting rollups in a base currency.

   ⚑ flag

Tax

1. finance leadSales tax, VAT, and GST are calculated on every invoice — natively or through a first-class tax integration — with the right registrations and rates

   weight 2 · round drawn
   Chargebee—none0/10

   No evidence in the pack shows Chargebee performing native or integrated sales tax/VAT/GST calculation, registrations, or rate management on invoices. The only related community comment explicitly states dissatisfaction with having to 'handle VAT' themselves, implying no seamless native solution, and no docs reference tax engines like Avalara/TaxJar or VAT/GST features.

   • [community] “So to be clear I absolutely love Chargebee! It's rich of features, very flexible and support is always friendly and ready to help. It's prob…”

   proof ↗⚑ flag

   Lago—none0/10

   No evidence pack item mentions tax calculation, VAT/GST/sales tax handling, tax registrations, tax rates, or integrations with tax engines (e.g., Avalara, Stripe Tax, TaxJar). The docs cover invoicing, plans, subscriptions, and dunning but never address tax compliance. Missing for 10: any mention of native tax calculation, tax rate/registration management, or a first-class tax-engine integration.

   ⚑ flag

Export

1. finance leadExport every subscription, invoice, and usage record — and port stored payment methods out to another provider — if I choose to leave

   weight 2 · round to Lago
   Chargebee~partialclaimed3/10

   Chargebee's docs confirm a generic Export API resource that creates export jobs and provides a download URL (chargebee-docs-6), implying some data-export capability, but there is no evidence specifying export of subscriptions, invoices, and usage records specifically, nor any mention of porting stored payment methods to another provider. missing for 10: explicit documentation of exporting subscriptions/invoices/usage data, any payment-method portability/export mechanism, and independent confirmation of successful migration workflows.

   • [claimed-docs] “Export resource represents an export job and contains the status of the job and the download URL, if the job is successfully completed.”

   proof ↗⚑ flag

   Lago~partialclaimed4/10

   Lago's API/OpenAPI schema and CSV export for analytics (lago-docs-15, lago-docs-16, lago-docs-1) mean a finance lead could theoretically pull subscription, invoice, and usage data out via API, and self-hosting (lago-docs-11) avoids lock-in of infrastructure. However, there is no documented bulk-export or migration tool, no explicit invoice/subscription/usage-record export workflow, and payment methods are held by external providers (Stripe etc.) rather than Lago itself, so 'porting stored payment methods' isn't addressed at all. Missing for 10: dedicated data-export/migration documentation, explicit payment-method portability mechanism, and any hands-on evidence of a successful migration off Lago.

   • [claimed-docs] “Build custom dashboards per customer, filter any way you want and export to CSV.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “You can download invoices or use the invoice object to create your own invoice template.”
   • [claimed-docs] “Docker is the easiest way to get started with the self-hosted version of Lago.”

   proof ↗⚑ flag

Import

1. ops userMigrate live subscriptions in from a spreadsheet or a competing platform — including PCI-compliant card data transfer — without customers noticing

   weight 3 · round drawn
   Chargebee—none0/10

   No evidence of any migration tooling, bulk import for live subscriptions, PCI-compliant card data transfer/tokenization import, or zero-downtime cutover capability is present in the evidence pack. Only general API/export/dunning/revrec docs are shown, none addressing subscription migration from spreadsheets or competitor platforms.

   ⚑ flag

   Lago—none0/10

   No evidence in the pack describes any migration tooling, bulk-import from spreadsheets, or PCI-compliant card-data transfer capability; the API/docs cover plan assignment and event ingestion but nothing about importing live subscriptions from another platform. Community comments even suggest Lago does not itself handle payment/PCI processing (lago-comm-8), reinforcing the absence of this capability.

   • [community] “"If I still have to pay the processing fees, what's the advantage here? Having to maintain my own payments stack (and PCI compliance) sounds…”

   proof ↗⚑ flag

1. ai-native userDo everything through the API that I can do in the UI

   weight 2 · round to Lago
   Chargebee~partialclaimed6/10

   Chargebee's REST API documentation covers a broad range of core UI-configurable features (subscriptions, usage events, exports, webhooks, entitlements) suggesting strong API/UI parity, and an API Explorer lets users test all documented endpoints. However, there's no explicit vendor claim of full UI-API parity, and features like AI drafting (chargebee-docs-14) or dunning setup are only described as UI features without confirmation they're also fully API-driven. Missing for 10: explicit parity statement, evidence that every UI action (e.g., AI drafting, dunning configuration) has an API equivalent, independent/hands-on confirmation of parity.

   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “This resource allows you to record usage events, which are essential for usage-based billing.”
   • [claimed-docs] “An item price is compatible with a plan-item price if their currencies are the same.”
   • [claimed-docs] “If webhooks are configured in Chargebee, events trigger those webhooks.”
   • [claimed-docs] “Export resource represents an export job and contains the status of the job and the download URL, if the job is successfully completed.”
   • [claimed-docs] “The API Explorer lets you make live API calls in a browser using a guided interface. Use it to test endpoints without writing code.”
   • [claimed-docs] “Chargebee provides you an efficient UI to allocate and manage the customer privileges using Features.”
   • [claimed-docs] “With Chargebee's Dunning feature, you can send email reminders to your customers asking them to update payment details and schedule payment …”
   • [claimed-docs] “Let Chargebee AI draft product families, plans, addons, charges, prices, features, and entitlements from a pricing page, document, or descri…”

   proof ↗⚑ flag

   Lago~partialprobed7/10

   Lago's docs show an API-first design: every endpoint has a generated CLI/code sample (lago-docs-1), a full OpenAPI 3.1 schema is available (lago-docs-16, lago-probe-3), and specific UI actions like usage-charge updates are explicitly available via both API and no-code UI (lago-docs-14). This implies strong API/UI parity but there is no explicit vendor claim or independent audit confirming 100% parity across every UI workflow (e.g., dunning, revenue analytics dashboards). missing for 10: an explicit 'everything in the UI is available via API' statement, and independent/community confirmation of full parity across all UI features.

   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “Both non-engineers and engineers can update usage charges via API or Lago’s no-code UI.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”

   proof ↗⚑ flag

2. ai-native userExport all of my data in open formats and leave

   weight 3 · round to Lago
   Chargebee~partialclaimed5/10

   Chargebee documents an Export resource that produces job status and a download URL (chargebee-docs-6), and its REST API returns JSON that could be used to pull data programmatically (chargebee-docs-1), giving a route to extract data in a machine-readable format. However there is no explicit documentation of bulk/full-account data export, no mention of standard open formats like CSV/JSON dumps of all account data, and no evidence of a clear 'leave with your data' workflow. missing for 10: explicit full-account bulk export documentation, confirmation of open/standard export formats (CSV/JSON) for all entities, and independent confirmation that this satisfies a complete data portability use case.

   • [claimed-docs] “Export resource represents an export job and contains the status of the job and the download URL, if the job is successfully completed.”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”

   proof ↗⚑ flag

   Lago~partialprobed6/10

   Lago's open-source/self-hosted nature (Docker deployment), open API (OpenAPI 3.1 schema, full REST reference), and CSV export of analytics/invoices give users real data-portability levers, and self-hosting means the underlying database is always in the user's own control. However there is no documented single 'export all account data' or account-closure/migration feature — only piecemeal exports (invoices, revenue analytics CSV, API access to individual objects). Missing for 10: a dedicated full-account data export/backup tool, explicit data-portability/migration documentation, and independent confirmation of a clean 'leave' workflow.

   • [claimed-docs] “Docker is the easiest way to get started with the self-hosted version of Lago.”
   • [claimed-docs] “You can download invoices or use the invoice object to create your own invoice template.”
   • [claimed-docs] “Build custom dashboards per customer, filter any way you want and export to CSV.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”

   proof ↗⚑ flag

3. ai-native userRead the product's source under an open license

   weight 2 · round to Lago
   Chargebee—none0/10

   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.)

   ⚑ flag

   Lago✓fullprobed7/10

   Lago's docs and GitHub explicitly describe it as an open-source billing platform with a public source repository (getlago/lago), and the llms.txt probe confirms this framing, satisfying the 'read source under open license' story. Missing for 10: explicit citation of the specific license name/file (e.g., AGPL) and independent verification of license terms beyond the 'open-source' label.

   • [probe] “PROBE llms.txt: HTTP 200 at https://getlago.com/llms.txt # Lago > Lago is the open-source billing platform for usage-based, subscription-ba…”
   • [github] “It starts the Lago version that matches this checkout, creates a disposable local organization, and prices three illustrative AI requests.”

   proof ↗⚑ flag

4. ai-native userSelf-host the core product

   weight 3 · round to Lago
   Chargebee—none0/10

   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.)

   ⚑ flag

   Lago✓fullcommunity8/10

   Docs explicitly cover self-hosting via Docker with a dedicated guide, and community comments confirm self-hosting is a real, used option ('self-hosting is not an issue', 'a self-hosted option fits our culture'). Missing for 10: independent hands-on report of running the full self-hosted stack in production, and more detail on self-hosted feature parity/limitations vs cloud.

   • [claimed-docs] “Docker is the easiest way to get started with the self-hosted version of Lago.”
   • [community] “"This looks great... Stripe is awesome but a self-hosted option fits our culture and philosophy a lot better. Will give it a blast when i ge…”
   • [community] “"Looks very interesting. We are small shop and self-hosting is not an issue. This neatly fits a requirement we've had for quite some time."”

   proof ↗⚑ flag

1. ai-native userChoose where my data is stored (region/residency)

   weight 2 · round to Lago
   Chargebee—none0/10

   No evidence in the pack discusses data residency, regional data storage options, or hosting location controls for Chargebee; the docs cover API usage, billing features, and MCP integration but nothing about choosing storage regions. Missing for 10: any mention of data residency/region selection, compliance certifications tied to geographic storage, or customer-facing controls for data location.

   ⚑ flag

   Lago~partialclaimed3/10

   Lago's self-hosted Docker deployment (lago-docs-11) implicitly lets a user control where their data resides by choosing their own infrastructure/region, but there is no explicit documentation of a region-selection feature, data-residency guarantees, or compliance certifications for the managed/cloud offering. missing for 10: explicit data-residency/region-selection feature docs, compliance certifications (e.g., GDPR/SOC2 region controls), managed-cloud region options.

   • [claimed-docs] “Docker is the easiest way to get started with the self-hosted version of Lago.”

   proof ↗⚑ flag

2. ai-native userPrevent my data from being used to train AI models

   weight 3 · round drawn
   Chargebee—none0/10

   No evidence in the pack addresses AI training data opt-out, data usage policies for model training, or any privacy controls specific to preventing data from being used to train AI models; evidence only covers billing API/MCP features unrelated to this privacy axis.

   ⚑ flag

   Lago—none0/10

   Lago is a billing platform; evidence covers billing, invoicing, self-hosting, and AI agent integrations (MCP server, dunning agent) but nowhere addresses data-use-for-AI-training opt-outs or privacy controls over model training. Missing for 10: any privacy policy statement, opt-out mechanism, or data-training exclusion clause.

   ⚑ flag

3. ai-native userControl data retention and deletion

   weight 2 · round drawn
   Chargebee—none0/10

   No evidence in the pack discusses data retention policies, deletion controls, or privacy/data lifecycle management for AI-native users; the pack covers billing APIs, MCP integrations, and dunning/revenue features but nothing about controlling or deleting stored data.

   ⚑ flag

   Lago—none0/10

   No evidence in the pack addresses data retention policies, deletion of customer/billing data, or GDPR-style controls; Lago's docs focus on billing, invoicing, and AI agent features but never mention data retention/deletion controls. Missing for 10: documentation on data retention windows, customer data deletion/export mechanisms, and any privacy/GDPR compliance controls.

   ⚑ flag

4. ai-native userOpt out of telemetry and usage tracking

   weight 2 · round drawn
   Chargebee—none0/10

   No evidence in the pack addresses telemetry/usage-tracking opt-out settings for Chargebee's product, MCP servers, or SDKs; this is an applicable privacy-posture axis for any SaaS/API product but no documentation or community evidence confirms such a control exists.

   ⚑ flag

   Lago—none0/10

   No evidence in the pack addresses telemetry, usage tracking, or opt-out settings for Lago itself (self-hosted or cloud). The docs cover billing features, AI agents, and self-hosting setup, but nothing about product analytics/telemetry collection or a way to disable it. Missing for 10: any mention of telemetry collection, an opt-out flag/config, or privacy documentation addressing data sent back to Lago.

   ⚑ flag

Analytics

1. founderI get MRR, churn, cohort retention, and LTV analytics computed from the billing source of truth, not a bolted-on BI export

   weight 2 · round drawn
   Chargebee—none0/10

   The evidence shows RevRec for ASC 606 compliance and generic data Export jobs, but nothing documents native MRR, churn, cohort retention, or LTV analytics computed within Chargebee itself — the only analytics-adjacent capability shown (Exports) is exactly the 'bolted-on export' pattern the story asks to avoid.

   • [claimed-docs] “Export resource represents an export job and contains the status of the job and the download URL, if the job is successfully completed.”
   • [claimed-docs] “RevRec is fully compliant with ASC 606 revenue recognition capabilities that cover your sales activities right from contract changes (renewa…”

   proof ↗⚑ flag

   Lago—none0/10

   Lago's docs mention a generic 'revenue-analytics' page with custom dashboards, filters, and CSV export (lago-docs-15), which is closer to the bolted-on BI export the story explicitly wants to avoid, not native MRR/churn/cohort-retention/LTV computation. No evidence pack item shows Lago computing these specific metrics natively from billing data, and a community question in lago-comm-3 explicitly asks whether Lago handles revenue recognition, left unanswered.

   • [claimed-docs] “Build custom dashboards per customer, filter any way you want and export to CSV.”
   • [community] “User describes building an internal billing tool after bad experiences with Recurly and ChargeBee for B2B/B2E usage-based billing, discounts…”

   proof ↗⚑ flag

Gl sync

1. finance leadInvoices, payments, and recognition entries sync to my general ledger or ERP — QuickBooks, Xero, NetSuite — so the close doesn't need CSV surgery

   weight 2 · round drawn
   Chargebee—none0/10

   The pack shows Chargebee has RevRec (ASC 606) and generic Export/API/webhook capabilities, but no evidence of any named QuickBooks, Xero, or NetSuite connector or GL/ERP sync workflow — missing for 10: named ERP/accounting integrations, sync mechanism details, and evidence that invoices/payments/recognition entries flow automatically without manual export/import.

   • [claimed-docs] “RevRec is fully compliant with ASC 606 revenue recognition capabilities that cover your sales activities right from contract changes (renewa…”
   • [claimed-docs] “Export resource represents an export job and contains the status of the job and the download URL, if the job is successfully completed.”
   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”

   proof ↗⚑ flag

   Lago—none0/10

   The evidence pack shows Lago's API, webhooks, invoicing, and OpenAPI schema, but there is no mention of any native or partner integration with QuickBooks, Xero, or NetSuite, nor any revenue-recognition export/sync to a GL/ERP system. Finance-lead close automation is a fair axis for a billing platform, but nothing in the pack demonstrates it.

   ⚑ flag

Revrec reports

1. finance leadRevenue is recognized on an ASC 606 / IFRS 15 basis — deferred revenue schedules and audit-ready recognition reports straight from billing data

   weight 3 · round to Chargebee
   Chargebee~partialclaimed6/10

   Chargebee's RevRec product is explicitly documented as ASC 606 compliant, covering contract changes like renewal/modification/cancellation, directly matching the finance-lead's revenue recognition need. However, evidence lacks detail on IFRS 15 specifically, deferred revenue schedule specifics, audit-ready report generation/export workflows, or independent/community corroboration of accuracy in practice. Missing for 10: IFRS 15 explicit coverage, deferred revenue schedule detail, audit trail/report examples, and third-party validation of RevRec accuracy.

   • [claimed-docs] “RevRec is fully compliant with ASC 606 revenue recognition capabilities that cover your sales activities right from contract changes (renewa…”

   proof ↗⚑ flag

   Lago—none0/10

   No evidence describes ASC 606/IFRS 15 deferred revenue schedules or audit-ready recognition reports; the closest material (revenue-analytics dashboards/CSV export) is generic reporting, not recognition accounting. A community comment even explicitly asks whether Lago handles revenue recognition, with no documented answer.

   • [claimed-docs] “Build custom dashboards per customer, filter any way you want and export to CSV.”
   • [community] “User describes building an internal billing tool after bad experiences with Recurly and ChargeBee for B2B/B2E usage-based billing, discounts…”

   proof ↗⚑ flag

Lifecycle api

1. developerCreate, update, pause, and cancel subscriptions through a documented API — the full lifecycle without touching a dashboard

   weight 3 · round drawn
   Chargebee~partialclaimed6/10

   Chargebee's REST API and Subscriptions resource are documented (chargebee-docs-1, chargebee-docs-4), and it supports form-encoded requests, JSON responses, and even a live API Explorer for testing endpoints (chargebee-docs-9), consistent with full CRUD-style subscription management. However, the evidence pack never explicitly cites 'create', 'update', 'pause', or 'cancel' subscription endpoints or their request/response schemas, so the full lifecycle claim is inferred rather than directly documented. missing for 10: explicit endpoint-level documentation for pause and cancel operations, and independent developer confirmation of using these endpoints end-to-end.

   • [claimed-docs] “It's a REST API that uses HTTP Basic Auth, accepts form-encoded requests, and returns JSON.”
   • [claimed-docs] “An item price is compatible with a plan-item price if their currencies are the same.”
   • [claimed-docs] “The API Explorer lets you make live API calls in a browser using a guided interface. Use it to test endpoints without writing code.”

   proof ↗⚑ flag

   Lago~partialprobed6/10

   Lago's docs cover API-driven subscription creation (assign-plan), upgrades/downgrades, and a full OpenAPI 3.1 schema with generated commands per endpoint, supporting a documented API lifecycle. However, the evidence pack never explicitly documents pause or cancel subscription endpoints, and most subscription guides (e.g., assign-plan) describe dashboard clicks rather than pure API-only flows. Missing for 10: explicit pause/cancel endpoint documentation, and hands-on/independent confirmation that the entire lifecycle can be done via API without the dashboard.

   • [claimed-docs] “Every endpoint in this reference has a generated command.”
   • [claimed-docs] “In the “Overview” tab, click “Add a plan” on the right”
   • [claimed-docs] “Customers have the flexibility to easily upgrade or downgrade their plan at any given moment.”
   • [claimed-docs] “Use Lago's OpenAPI 3.1 schema to generate clients, inspect typed operations, or expose billing actions to an AI agent.”
   • [probe] “PROBE openapi: HTTP 200 at https://getlago.com/openapi.json — contains "openapi" key”

   proof ↗⚑ flag

Lifecycle events

1. developerEvery subscription state change — renewal, upgrade, payment failure, cancellation — emits a webhook or event my systems can act on reliably

   weight 2 · round to Chargebee
   Chargebee✓fullcommunity7/10

   Chargebee's Events API triggers configured webhooks (chargebee-docs-5), and community evidence confirms multiple webhook URLs can be configured for different environments, corroborating reliable event delivery (chargebee-comm-2). Missing for 10: explicit documentation of the full set of lifecycle event types (renewal, upgrade, payment failure, cancellation) and delivery-retry/reliability guarantees for webhook events.

   • [claimed-docs] “If webhooks are configured in Chargebee, events trigger those webhooks.”
   • [community] “With Chargebee I can add multiple webhook URLs, so all events get sent to both dev and staging and each environment only cares about its own…”

   proof ↗⚑ flag

   Lago~partialclaimed6/10

   Lago documents a webhook system with signing (HMAC) and explicitly treats 'webhook event types' as a versioned API surface, implying granular event types exist, but the evidence never enumerates specific subscription lifecycle events (renewal, upgrade, payment failure, cancellation) or demonstrates delivery reliability guarantees like retries/idempotency for webhooks specifically. missing for 10: explicit list/documentation of subscription lifecycle event types, evidence of webhook delivery reliability (retries, dead-letter, idempotency) beyond event-ingestion dedup which is a different feature.

   • [claimed-docs] “The second is `HMAC`, which features a shorter payload header and no size restrictions.”
   • [claimed-docs] “Clients should tolerate additive changes within v1, including new response fields, enum values, endpoints, optional request fields, and webh…”

   proof ↗⚑ flag

Self serve

1. founderMy customers can upgrade, downgrade, or cancel themselves through a hosted portal or embeddable components, without a support ticket

   weight 2 · round drawn
   Chargebee—none0/10

   The evidence pack covers Chargebee's API, dunning, entitlements admin UI, and MCP/AI tooling, but nothing describes a hosted customer portal or embeddable self-service components letting end customers upgrade/downgrade/cancel on their own. missing for 10: hosted customer portal documentation, embeddable self-service widget docs, evidence of ticket-free self-service flow.

   ⚑ flag

   Lago—none0/10

   Evidence confirms upgrade/downgrade functionality exists (lago-docs-4, lago-docs-3) but only via API or Lago's internal admin UI — there is no mention anywhere in the pack of a customer-facing hosted portal or embeddable self-serve components that end customers could use directly. Community comments even question whether Lago handles full self-serve lifecycle (lago-comm-3), and nothing confirms a portal exists.

   • [claimed-docs] “Customers have the flexibility to easily upgrade or downgrade their plan at any given moment.”
   • [claimed-docs] “In the “Overview” tab, click “Add a plan” on the right”
   • [community] “User describes building an internal billing tool after bad experiences with Recurly and ChargeBee for B2B/B2E usage-based billing, discounts…”

   proof ↗⚑ flag

Trials promos

1. ops userRun free trials, coupons, and promotion codes with automatic conversion to paid — including trial-end notifications required by card-network rules

   weight 2 · round drawn
   Chargebee—none0/10

   The evidence pack covers dunning/smart retry, entitlements, RevRec, and multi-frequency billing, but contains no mention of free trials, coupon/promotion code functionality, automatic trial-to-paid conversion, or trial-end notification requirements — the specific subscription-lifecycle capabilities named in the story are entirely undocumented here.

   ⚑ flag

   Lago—none0/10

   No evidence pack items mention free trials, coupons, promotion codes, automatic trial-to-paid conversion, or trial-end notifications required by card-network rules; evidence covers plans, subscriptions, invoicing, webhooks, and AI agents but not this trial/coupon lifecycle. missing for 10: docs on free trials, coupons/promo codes, automatic conversion to paid, and trial-end notification compliance.

   ⚑ flag

Event ingestion

1. developerStream high-volume usage events — API calls, tokens, compute-seconds — into a metering pipeline with idempotent ingestion and backdating support

   weight 3 · round to Lago
   Chargebee~partialclaimed3/10

   Chargebee's API docs confirm a usage_events resource for recording usage-based billing events, but the evidence pack contains no detail on idempotent ingestion, backdating, or high-volume streaming capability. Missing for 10: idempotency key support, backdating/late event handling, batch/streaming ingestion guarantees, and any performance/scale documentation.

   • [claimed-docs] “This resource allows you to record usage events, which are essential for usage-based billing.”

   proof ↗⚑ flag

   Lago~partialclaimed5/10

   Lago's docs confirm idempotent event ingestion (dedup on repeated event IDs) and a usage-metering pipeline supporting API-driven event updates, which covers the core ingestion/idempotency claims. However, there is no explicit evidence of backdating support (e.g., specifying a past timestamp for events) or of high-volume streaming architecture/throughput guarantees for API calls, tokens, or compute-seconds use cases. missing for 10: explicit backdating/timestamp-override documentation, evidence of high-volume/streaming ingestion architecture or throughput benchmarks, independent verification of dedup behavior at scale.

   • [claimed-docs] “Lago uses it for deduplication: if the same event arrives twice, it is billed once.”
   • [claimed-docs] “Both non-engineers and engineers can update usage charges via API or Lago’s no-code UI.”

   proof ↗⚑ flag

Pricing models

1. finance leadPrice with tiered, volume, graduated, per-unit, and hybrid subscription-plus-usage models — including credits, commitments, and overage — without custom code

   weight 3 · round to Lago
   Chargebee~partialclaimed3/10

   Evidence confirms usage-event recording for usage-based billing and multi-frequency subscription billing, which supports basic usage-plus-subscription models, but there is no documentation of tiered, volume, graduated, per-unit, or hybrid pricing structures, nor of credits, commitments, or overage handling. missing for 10: explicit tiered/volume/graduated/per-unit pricing model docs, credit/commitment/overage handling documentation, no-code configuration evidence for these pricing types.

   • [claimed-docs] “This resource allows you to record usage events, which are essential for usage-based billing.”
   • [claimed-docs] “Multi-frequency billing: Subscriptions are billed according to the billing period of each recurring item price within the subscription, whet…”
   • [claimed-docs] “An item price is compatible with a plan-item price if their currencies are the same.”

   proof ↗⚑ flag

   Lago~partialprobed5/10

   Lago's own positioning ('usage-based, subscription-based, and hybrid pricing models') and the plan-model/usage-metering docs confirm broad subscription+usage flexibility, and community reviews praise it for handling complex billing without custom code, but the evidence pack never explicitly documents tiered, volume, graduated, or per-unit charge models, nor credits, commitments, or overage handling as distinct configurable features. missing for 10: explicit docs on tiered/volume/graduated/per-unit charge model configuration, credits and commitment mechanics, overage billing rules, and independent confirmation these work without custom code.

   • [probe] “PROBE llms.txt: HTTP 200 at https://getlago.com/llms.txt # Lago > Lago is the open-source billing platform for usage-based, subscription-ba…”
   • [claimed-docs] “Yearly: subscription fees are billed on a yearly basis and charges can be billed monthly or annually.”
   • [claimed-docs] “Both non-engineers and engineers can update usage charges via API or Lago’s no-code UI.”
   • [claimed-docs] “Spot overdue, partially paid or unpaid invoices with filters to never lose track of what you’re owed.”
   • [community] “"We decided to use Lago with our SaaS. Billing is super complicated and Lago handles everything we need out of the box. Building this stuff …”
   • [community] “"Lago is awesome. We'll likely use Lago at my startup (metlo) when we implement full self-serve usage-based billing."”

   proof ↗⚑ flag

Usage visibility

1. ops userCustomers and my team can see current-cycle usage and accrued cost in real time, with thresholds and alerts before an invoice surprises anyone

   weight 2 · round to Lago
   Chargebee~partialclaimed3/10

   Chargebee's usage_events API supports recording usage for usage-based billing and webhooks can propagate events, but there is no documented dashboard for real-time cycle usage/cost visibility to customers and teams, nor any explicit threshold/alert feature for pre-invoice warnings. Missing for 10: real-time usage/cost dashboard for customers, configurable usage thresholds, proactive alerting before invoicing.

   • [claimed-docs] “This resource allows you to record usage events, which are essential for usage-based billing.”
   • [claimed-docs] “If webhooks are configured in Chargebee, events trigger those webhooks.”

   proof ↗⚑ flag

   Lago~partialclaimed5/10

   Lago documents usage-metering and customer-facing analytics dashboards (lago-docs-14, lago-docs-15) and webhooks for event notifications (lago-docs-7), which could support real-time usage visibility, but there is no explicit documentation of configurable usage thresholds or proactive alerting before invoice generation. missing for 10: explicit threshold/alert configuration feature, evidence of real-time cost accrual view distinct from dashboards, confirmation that customers (not just ops) see live usage/cost.

   • [claimed-docs] “Both non-engineers and engineers can update usage charges via API or Lago’s no-code UI.”
   • [claimed-docs] “Build custom dashboards per customer, filter any way you want and export to CSV.”
   • [claimed-docs] “The second is `HMAC`, which features a shorter payload header and no size restrictions.”

   proof ↗⚑ flag