Stripe vs Checkout.com

Agent access

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

   weight 2 · round to Stripe
   Stripe✓fullprobed9/10

   Stripe hosts a live llms.txt (probe confirms HTTP 200 with agent-oriented guidance), a full .md-suffixed docs mirror (docs.stripe.com/*.md), a machine-readable skills index at /.well-known/skills/index.json, and dedicated agent-oriented docs (agents.md, mcp.md, directory.md) explicitly designed for pointing agents at. This is independently verified by probes, not just vendor claims. missing for 10: no independent third-party (non-Stripe) report of an agent successfully consuming llms.txt end-to-end in production.

   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.stripe.com/llms.txt # Stripe Documentation When installing Stripe packages, always check the npm r…”
   • [probe] “PROBE docs-md: HTTP 200 at https://docs.stripe.com/payments.md # Payments Use Stripe to start accepting payments. ## Get started Integrat…”
   • [claimed-docs] “Stripe publishes a machine-readable agent-skills catalog: "You can use curl to fetch the index of available skills, what they do, and their …”
   • [probe] “PROBE skills-index (2026-09-14): GET https://docs.stripe.com/.well-known/skills/index.json returns HTTP 200 with 8 Stripe-maintained agent s…”
   • [claimed-docs] “Build your integration with agent-first developer tools, charge your users for AI features with usage-based billing, and let agents transact…”
   • [claimed-docs] “The Stripe Model Context Protocol (MCP) server provides tools that AI agents can use to interact with the Stripe API and search Stripe's kno…”

   proof ↗⚑ flag

   Checkout.com✓fullprobed8/10

   Checkout.com hosts a live llms.txt at both www.checkout.com/llms.txt and docs.checkout.com/llms.txt, confirmed via probe returning HTTP 200 with structured content, plus multiple docs entries derived from that file listing agent-readable summaries of products/APIs. This directly satisfies pointing an agent at llms.txt. missing for 10: independent third-party confirmation of an agent successfully consuming the file end-to-end, and a docs.md/markdown mirror (docs.md returned 404).

   • [probe] “PROBE llms.txt: HTTP 200 at https://www.checkout.com/llms.txt # Checkout.com: High-performance global payments platform for enterprise and s…”
   • [claimed-docs] “Unified Payments API: Connect to all payment methods through a single API integration.”
   • [claimed-docs] “Request a payment using a token, a network token, a payment instrument, or the full card details.”
   • [claimed-docs] “Use your Checkout Business Account to efficiently manage your funds across our product suite.”
   • [probe] “PROBE docs-md: HTTP 404 at https://www.checkout.com/docs.md”

   proof ↗⚑ flag

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

   weight 2 · round to Stripe
   Stripe✓fullclaimed9/10

   Stripe's CLI and Workbench Shell provide scriptable, non-interactive access (stripe sandbox create, stripe products create, stripe listen, agent setup) usable in CI pipelines, backed by idempotency keys, restricted API keys, and a documented sandbox that requires no account registration — all designed for headless/automated use including by coding agents. missing for 10: no explicit first-party CI example (e.g. GitHub Actions workflow) or independent hands-on report of running the CLI in a CI pipeline.

   • [claimed-docs] “Coding agents should install the Stripe CLI (npm i -g @stripe/cli) and run the command stripe sandbox create --help to provision an anonymou…”
   • [claimed-docs] “Build, test, and manage your Stripe integration from the command line.”
   • [claimed-docs] “Workbench (Dashboard > Developers > Workbench) ships an interactive Shell and API Explorer: "Shell is a command line interface within Workbe…”
   • [claimed-docs] “The CLI landing page frames the command line as "Stripe's suite of agent-ready tooling", with "Agent setup: Automatically configure your age…”
   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”
   • [claimed-docs] “Restricted API keys are Stripe's documented least-privilege credential for agents: "When you create a RAK in the Stripe Dashboard, you selec…”
   • [claimed-docs] “A sandbox is an isolated test environment. You can use your sandbox to test Stripe functionality in your account, and experiment with new fe…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed7/10

   Checkout.com is API-first with SDKs (e.g. Node.js), OAuth2 client-credentials auth, idempotent retries, webhooks, and a full sandbox/test-account environment with test cards — all of which support scripted, headless use in CI/test pipelines. However there is no dedicated CLI or explicit CI/CD pipeline documentation/tutorials for automation. Missing for 10: dedicated CLI tool, explicit CI/CD integration guide or example pipeline.

   • [claimed-docs] “Idempotent retries are first-class: "The following endpoints from our API support idempotent requests, so you can safely retry them without …”
   • [claimed-docs] “Two documented server-side authentication forms — "Access keys (OAuth 2.0)" (client-credentials flow) and "Secret keys" — plus public keys f…”
   • [claimed-docs] “A dedicated test-account environment with per-scenario testing docs: payments testing, test cards, disputes testing, fraud-detection testing…”
   • [github] “The official Node.js SDK for Checkout.com payment gateway.”
   • [claimed-docs] “Test cards”

   proof ↗⚑ flag

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

   weight 3 · round drawn
   Stripe✓fullprobed9/10

   Stripe ships a well-documented, independently probed official hosted MCP server (mcp.stripe.com) exposing API tools, analytics, and knowledge-base search, with OAuth-based auth/governance, human-confirmation for write actions, and connected-account support; a live probe confirms the OAuth-gated endpoint is real and functioning. Missing for 10: no independent third-party hands-on review of actual agent usage beyond Stripe's own docs/probes.

   • [claimed-docs] “The Stripe Model Context Protocol (MCP) server provides tools that AI agents can use to interact with the Stripe API and search Stripe's kno…”
   • [claimed-docs] “Let your AI agents interact with the Stripe API by using our MCP server.”
   • [probe] “official MCP server documented at https://docs.stripe.com/mcp”
   • [probe] “PROBE mcp-oauth (2026-09-14): keyless JSON-RPC initialize to https://mcp.stripe.com answers HTTP 401 {"error":"Unauthorized. See https://doc…”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “Docs, "Confirm actions by agents acting on your behalf": "To prevent agents from making mistakes, Stripe requires human confirmation before …”
   • [claimed-docs] “MCP access is governed like any credential: OAuth per the MCP spec ("OAuth lets you grant and revoke access without sharing an API key with …”
   • [claimed-docs] “Docs, "Use MCP with connected accounts": a Connect platform "can also use MCP to retrieve or update resources for your connected accounts...…”

   proof ↗⚑ flag

   Checkout.com✓fullprobed9/10

   Checkout.com documents an official, hosted MCP server (sandbox and production endpoints, OAuth via Dashboard account) with concrete setup instructions (e.g., adding it to Claude Code) and documented capabilities like querying payments, issuing refunds, managing payment links, and knowledge-base search — directly fulfilling the story. Missing for 10: independent/third-party corroboration of real-world usage beyond vendor docs, and it's explicitly marked Beta.

   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”
   • [claimed-docs] “The Checkout.com Model Context Protocol (MCP) Server enables you to search through our knowledge base and manage payment operations directly…”
   • [probe] “official MCP server documented at https://www.checkout.com/docs/developer-resources/checkout-com-mcp-server”

   proof ↗⚑ flag

4. ai-native userUse an official CLI

   weight 2 · round to Stripe
   Stripe✓fullprobed9/10

   Stripe ships a well-documented official CLI (docs.stripe.com/stripe-cli) for building, testing, and managing integrations, explicitly positioned as 'agent-ready tooling' with commands like `stripe agent setup` (auto-configures Claude Code/Codex/Cursor), `stripe sandbox create` for keyless agent provisioning, `stripe directory search`, and `stripe projects init`. This is corroborated by an independent probe confirming the docs page exists and by Workbench's browser-based Shell mirroring CLI commands. Missing for 10: independent hands-on/community review specifically validating CLI usability for AI-native workflows beyond Stripe's own docs.

   • [claimed-docs] “Build, test, and manage your Stripe integration from the command line.”
   • [claimed-docs] “Workbench provides powerful tools that help you build, test, and debug your Stripe integration.”
   • [probe] “official CLI documented at https://docs.stripe.com/stripe-cli”
   • [claimed-docs] “The CLI landing page frames the command line as "Stripe's suite of agent-ready tooling", with "Agent setup: Automatically configure your age…”
   • [claimed-docs] “Official agent plugins bundle MCP + skills with auto-update: "Stripe's agent plugins give your AI agent access to Stripe developer tools, li…”
   • [claimed-docs] “Coding agents should install the Stripe CLI (npm i -g @stripe/cli) and run the command stripe sandbox create --help to provision an anonymou…”
   • [claimed-docs] “Stripe Projects "provisions and manages third-party services (such as hosting, databases, auth, AI, and observability) from the terminal. Ru…”
   • [claimed-docs] “"Stripe Directory helps developers and AI agents find the best external providers for a task and follow the best supported path to provision…”

   proof ↗⚑ flag

   Checkout.com—none0/10

   Evidence shows SDKs (Node.js), a Dashboard, an MCP server, and agentic-commerce protocol integrations, but no mention of an official CLI tool for developers. Payment platforms commonly ship CLIs (e.g., for testing webhooks, managing keys), so the axis applies, but no such capability is documented anywhere in this evidence pack.

   • [github] “The official Node.js SDK for Checkout.com payment gateway.”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [probe] “PROBE docs-md: HTTP 404 at https://www.checkout.com/docs.md”
   • [probe] “PROBE openapi: all candidate paths 404 (https://www.checkout.com/openapi.json, https://www.checkout.com/swagger.json, https://www.checkout.c…”

   proof ↗⚑ flag

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

   weight 3 · round drawn
   Stripe✓fullprobed9/10

   Stripe exposes a comprehensive, well-documented REST API with idempotency, restricted keys, versioning, sandboxes, webhooks, OpenAPI-generated SDKs, and CLI/Workbench tooling, and this is independently corroborated by a probe confirming the API and llms.txt are live and reachable. missing for 10: no independent third-party benchmark of raw API completeness beyond docs/probes, and OpenAPI spec discovery via well-known paths returned 404s per stripe-probe-3.

   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”
   • [claimed-docs] “use an idempotency key... you can safely repeat the request without risk of creating a second object or performing the update twice”
   • [claimed-docs] “Stripe uses API keys to authenticate requests from your integration and determine which Stripe resources it can access.”
   • [claimed-docs] “A restricted API key (RAK) is an API key ... a RAK can do only what you give it permission to do.”
   • [claimed-docs] “You can safely upgrade to a new monthly release without breaking any existing code.”
   • [claimed-docs] “Each monthly release includes only backward-compatible changes... You can safely upgrade to a new monthly release without breaking any exist…”
   • [claimed-docs] “Use these files to generate SDKs or client libraries that support Stripe's API.”
   • [probe] “PROBE docs-md: HTTP 200 at https://docs.stripe.com/payments.md # Payments Use Stripe to start accepting payments. ## Get started Integrat…”
   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.stripe.com/llms.txt # Stripe Documentation When installing Stripe packages, always check the npm r…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.stripe.com/openapi.json, https://docs.stripe.com/swagger.json, https://docs.stripe.com/…”
   • [claimed-docs] “Build, test, and manage your Stripe integration from the command line.”
   • [claimed-docs] “Workbench (Dashboard > Developers > Workbench) ships an interactive Shell and API Explorer: "Shell is a command line interface within Workbe…”

   proof ↗⚑ flag

   Checkout.com✓fullprobed9/10

   Checkout.com documents a comprehensive public REST/Unified Payments API with OAuth2/API-key auth, idempotency, rate limits, webhooks, SDKs (Node.js), and a full testing sandbox, plus explicit agentic-commerce integrations (ChatGPT, Google AI) and an official MCP server for AI-agent operation — strongly supporting AI-native programmatic control. Missing for 10: a discoverable OpenAPI/swagger spec (probe found 404s) and independent third-party corroboration beyond vendor docs.

   • [claimed-docs] “Unified Payments API: Connect to all payment methods through a single API integration.”
   • [claimed-docs] “Idempotent retries are first-class: "The following endpoints from our API support idempotent requests, so you can safely retry them without …”
   • [claimed-docs] “Two documented server-side authentication forms — "Access keys (OAuth 2.0)" (client-credentials flow) and "Secret keys" — plus public keys f…”
   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “A dedicated test-account environment with per-scenario testing docs: payments testing, test cards, disputes testing, fraud-detection testing…”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”
   • [claimed-docs] “"With agentic commerce, your customers can discover and pay for products through an AI chat interface, instead of visiting your website. The…”
   • [github] “The official Node.js SDK for Checkout.com payment gateway.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://www.checkout.com/openapi.json, https://www.checkout.com/swagger.json, https://www.checkout.c…”

   proof ↗⚑ flag

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

   weight 2 · round to Stripe
   Stripe✓fullprobed9/10

   Stripe explicitly documents restricted API keys (RAKs) as the recommended least-privilege credential for AI agents—developers select per-resource Read/Write/None permissions—and pairs this with MCP OAuth sessions (per-client revocation, connected-account scoping via Stripe-Account header, human confirmation for write actions like refunds, and Workbench tool-call logs), giving concrete scoped-credential issuance and governance for agents. This is corroborated by a live probe confirming the OAuth-gated MCP endpoint. Missing for 10: independent/hands-on developer confirmation of the RAK issuance workflow in practice beyond first-party docs.

   • [claimed-docs] “A restricted API key (RAK) is an API key ... a RAK can do only what you give it permission to do.”
   • [claimed-docs] “Restricted API keys are Stripe's documented least-privilege credential for agents: "When you create a RAK in the Stripe Dashboard, you selec…”
   • [claimed-docs] “MCP access is governed like any credential: OAuth per the MCP spec ("OAuth lets you grant and revoke access without sharing an API key with …”
   • [claimed-docs] “Docs, "Use MCP with connected accounts": a Connect platform "can also use MCP to retrieve or update resources for your connected accounts...…”
   • [claimed-docs] “Docs, "Confirm actions by agents acting on your behalf": "To prevent agents from making mistakes, Stripe requires human confirmation before …”
   • [probe] “PROBE mcp-oauth (2026-09-14): keyless JSON-RPC initialize to https://mcp.stripe.com answers HTTP 401 {"error":"Unauthorized. See https://doc…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed5/10

   Checkout.com documents OAuth 2.0 client-credentials flow and 'access keys' alongside secret keys, letting integrations obtain scoped, non-static credentials per processing channel — a real least-privilege mechanism. However, the MCP server (the actual agent-facing surface) authenticates via a full Dashboard account login rather than a dedicated scoped/agent-specific credential, so there's no documented way to mint a narrowly-scoped key specifically for an autonomous agent. Missing for 10: explicit agent-scoped OAuth scopes/roles for the MCP server, granular permission docs tied to agent use cases, and independent confirmation that access-key scopes can be limited to agent-safe operations only.

   • [claimed-docs] “Two documented server-side authentication forms — "Access keys (OAuth 2.0)" (client-credentials flow) and "Secret keys" — plus public keys f…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”

   proof ↗⚑ flag

7. ai-native userBuild against official SDKs

   weight 2 · round to Stripe
   Stripe✓fullprobed9/10

   Stripe publishes official SDKs/client libraries generated from OpenAPI specs, plus AI-native additions like @stripe/agent-toolkit, @stripe/ai-sdk, token-meter, a hosted OAuth-gated MCP server, agent plugins for Claude/Codex/Cursor, and a machine-readable skills catalog — strongly evidencing an AI-native developer can build against official SDKs. missing for 10: independent (non-Stripe) hands-on developer reports specifically validating the newer agent-toolkit/ai-sdk packages in production.

   • [claimed-docs] “Use these files to generate SDKs or client libraries that support Stripe's API.”
   • [claimed-docs] “The stripe/ai repo (github.com/stripe/agent-toolkit now redirects here) is "the one-stop shop for building AI-powered products and businesse…”
   • [claimed-docs] “Official agent plugins bundle MCP + skills with auto-update: "Stripe's agent plugins give your AI agent access to Stripe developer tools, li…”
   • [claimed-docs] “Stripe publishes a machine-readable agent-skills catalog: "You can use curl to fetch the index of available skills, what they do, and their …”
   • [probe] “PROBE mcp-oauth (2026-09-14): keyless JSON-RPC initialize to https://mcp.stripe.com answers HTTP 401 {"error":"Unauthorized. See https://doc…”
   • [probe] “PROBE skills-index (2026-09-14): GET https://docs.stripe.com/.well-known/skills/index.json returns HTTP 200 with 8 Stripe-maintained agent s…”
   • [claimed-docs] “The Stripe Model Context Protocol (MCP) server provides tools that AI agents can use to interact with the Stripe API and search Stripe's kno…”
   • [github] “a collection of SDKs to help you integrate Stripe with LLMs and agent frameworks”

   proof ↗⚑ flag

   Checkout.com~partialprobed6/10

   Checkout.com documents an official Node.js SDK on GitHub, a native Mobile SDK for iOS/Android, plugins for ecommerce platforms, and machine-readable docs (llms.txt) plus an MCP server that let AI agents build against its API. However, evidence only names one language SDK explicitly and no broader SDK catalog (e.g., Python, PHP, Java, .NET) or independent hands-on validation is shown. Missing for 10: multi-language official SDK coverage, independent/community corroboration of SDK quality, and a public OpenAPI spec (probes show openapi.json 404).

   • [github] “The official Node.js SDK for Checkout.com payment gateway.”
   • [claimed-docs] “Mobile SDK: Build native mobile checkout experiences for iOS and Android.”
   • [claimed-docs] “Plugins: Integrate Checkout.com into existing ecommerce platforms via pre-built plugins.”
   • [claimed-docs] “Unified Payments API: Connect to all payment methods through a single API integration.”
   • [probe] “PROBE llms.txt: HTTP 200 at https://www.checkout.com/llms.txt # Checkout.com: High-performance global payments platform for enterprise and s…”
   • [claimed-docs] “The Checkout.com Model Context Protocol (MCP) Server enables you to search through our knowledge base and manage payment operations directly…”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”

   proof ↗⚑ flag

8. ai-native userSubscribe to events via webhooks

   weight 2 · round to Checkout.com
   Stripe✓fullclaimed8/10

   Stripe supports registering HTTPS webhook endpoints that push real-time events (payment success, disputes, subscription renewals, etc.), and this is directly usable by AI agents via the MCP server's stripe_api_write tool (webhook endpoint creation is explicitly listed as a supported MCP method) and via event-destinations delivering to multiple destination types. missing for 10: no hands-on/independent example of an agent actually subscribing to and consuming webhook events end-to-end, and no dedicated agent-specific webhook subscription API distinct from the standard developer webhook flow.

   • [claimed-docs] “Receiving webhook events helps you respond to asynchronous events, such as when a customer's bank confirms a payment, a customer disputes a …”
   • [claimed-docs] “Stripe pushes real-time data to it when events happen in your Stripe account.”
   • [claimed-docs] “You can create an HTTPS webhook endpoint to receive events.”
   • [claimed-docs] “After you register a webhook endpoint, Stripe pushes real-time data to it when events happen in your Stripe account.”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “Event destinations deliver real-time events "across multiple destination types, including webhook endpoints, Amazon EventBridge, and Azure E…”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed9/10

   Checkout.com has robust, well-documented webhooks: event notifications with HMAC signature verification, configuration via Dashboard or API, resending webhooks, and integration with Amazon EventBridge for event streaming — directly enabling programmatic/agentic subscription to payment lifecycle events. Missing for 10: independent/hands-on corroboration beyond first-party docs.

   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “Receive webhooks”
   • [claimed-docs] “Resend webhooks”
   • [claimed-docs] “Manage webhooks in the Dashboard”

   proof ↗⚑ flag

Agentic features

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

   weight 2 · round to Stripe
   Stripe~partialclaimed6/10

   Stripe supports AI-generated insights primarily through its MCP `stripe_analytics`/`stripe_report` tools, which let an AI agent query metrics like MRR, churn rate, and run custom SQL against reporting tables, and Radar applies AI algorithms to score fraud risk in real time. However, these insights are largely agent-mediated (via MCP) rather than a native in-Dashboard AI assistant proactively surfacing suggestions to the user. Missing for 10: evidence of a built-in conversational/insights UI inside the Dashboard itself (not just via external agent+MCP), broader proactive 'suggestions' beyond analytics queries, and independent/hands-on validation of insight quality.

   • [claimed-docs] “You can use an AI agent to analyze aggregated data across your Stripe account ... query metrics such as monthly recurring revenue (MRR)”
   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”
   • [claimed-docs] “Use Radar to evaluate transactions, accounts, and customers in real time, using AI algorithms to assess the risk of fraud.”
   • [claimed-docs] “Recovery analytics: Analyze subscription payment failure rates, recovery rates, and recent failed payments for top customers.”

   proof ↗⚑ flag

   Checkout.com~partialclaimed3/10

   Checkout.com references a general 'AI-driven solution' to optimize payment acceptance rates and mentions 'advanced analytics' on transaction-level data, which gestures at AI working on the merchant's own data, but there is no documented dashboard AI-insights/copilot feature that surfaces suggestions or narrative insights to the user. Missing for 10: a named in-product AI insights/analytics feature, evidence of generated recommendations shown to users, and any hands-on/independent corroboration beyond marketing copy.

   • [claimed-docs] “Automatically optimize payments to increase your acceptance rate with our powerful AI-driven solution.”
   • [claimed-docs] “Transaction-level data you can use to do more. Go beyond approvals and declines and experience our advanced analytics.”

   proof ↗⚑ flag

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

   weight 2 · round to Stripe
   Stripe~partialclaimed7/10

   Stripe ships substantial infrastructure for background automation — webhooks that push real-time events for asynchronous processing, Billing's automated recurring payments/renewals, Smart Retries and revenue-recovery automation that requires no code, Radar's automatic fraud scoring, and an MCP server plus agent plugins that let AI agents act on Stripe data continuously. However, Stripe explicitly gates autonomous execution: 'Stripe requires human confirmation before it takes certain stripe_api_write actions, such as refunds and outbound payments,' meaning fully unattended agentic automation is capped for sensitive operations, and there's no independent/hands-on evidence of long-running autonomous background agent workflows in production. Missing for 10: evidence of fully unattended agent automations without human-in-the-loop gating, and independent corroboration of real-world background agent runs.

   • [claimed-docs] “Receiving webhook events helps you respond to asynchronous events, such as when a customer's bank confirms a payment, a customer disputes a …”
   • [claimed-docs] “Smart Retries: Automatically retry failed payments to prevent involuntary churn due to temporary issues.”
   • [claimed-docs] “Recovery analytics: Analyze subscription payment failure rates, recovery rates, and recent failed payments for top customers.”
   • [claimed-docs] “Use Radar to evaluate transactions, accounts, and customers in real time, using AI algorithms to assess the risk of fraud.”
   • [claimed-docs] “Docs, "Confirm actions by agents acting on your behalf": "To prevent agents from making mistakes, Stripe requires human confirmation before …”
   • [claimed-docs] “Official agent plugins bundle MCP + skills with auto-update: "Stripe's agent plugins give your AI agent access to Stripe developer tools, li…”
   • [claimed-docs] “"Machine payments let agents pay for APIs and services programmatically. Your server returns a payment challenge, the agent presents a valid…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed4/10

   Checkout.com offers building blocks that support background automation — webhooks trigger server-side actions on payment lifecycle events, and the MCP server lets an AI agent query statuses, issue refunds, and manage payment links — plus agentic commerce protocols (ChatGPT, Google AI) letting an AI complete transactions on a user's behalf. However, none of this constitutes a dedicated 'set up an automation that runs autonomously in the background' capability (e.g., scheduled jobs, persistent agent workflows); the MCP server is described as an IDE-session tool, not an unattended background process. Missing for 10: a documented scheduler/background-job or persistent autonomous-agent workflow feature, and independent evidence of such automations running unattended.

   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “"With agentic commerce, your customers can discover and pay for products through an AI chat interface, instead of visiting your website. The…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”

   proof ↗⚑ flag

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

   weight 3 · round drawn
   Stripe—none0/10

   The evidence pack shows Stripe exposing an MCP server, CLI, SDKs, and agent plugins so *external* AI agents (Claude, Codex, Cursor, etc.) can act on Stripe's behalf, and it describes AI-driven backend features (Radar fraud scoring, smart retries). But none of the evidence describes a built-in, in-product AI assistant/chat feature embedded in the Stripe Dashboard itself that a user converses with to delegate tasks — the 'analyze with AI' and analytics tooling are also delivered via the external MCP interface rather than a native assistant UI.

   • [claimed-docs] “You can use an AI agent to analyze aggregated data across your Stripe account ... query metrics such as monthly recurring revenue (MRR)”
   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”
   • [claimed-docs] “Use Radar to evaluate transactions, accounts, and customers in real time, using AI algorithms to assess the risk of fraud.”

   proof ↗⚑ flag

   Checkout.com—none0/10

   Checkout.com offers an MCP server so external AI agents/IDEs can query its API, and supports agentic-commerce protocols so third-party AI agents (ChatGPT, Google AI) can transact with merchants — but there is no evidence of a built-in AI assistant embedded inside the Checkout.com product/dashboard itself that a user can delegate tasks to. The MCP 'Guide tool' and AI-driven acceptance optimization are external-agent or backend features, not an in-product assistant.

   • [claimed-docs] “The Checkout.com Model Context Protocol (MCP) Server enables you to search through our knowledge base and manage payment operations directly…”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “"With agentic commerce, your customers can discover and pay for products through an AI chat interface, instead of visiting your website. The…”
   • [claimed-docs] “Automatically optimize payments to increase your acceptance rate with our powerful AI-driven solution.”

   proof ↗⚑ flag

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

   weight 2 · round to Checkout.com
   Stripe~partialprobed7/10

   Stripe exposes an MCP server plus Directory/Analytics tools that let AI agents perform natural-language-driven actions (search products, read/write API resources, query MRR, run implementation planner) confirmed live via probe, and CLI/agent-plugin tooling lets coding agents set up sandboxes and integrations via NL prompts. However, this is agent-mediated (a separate AI agent translates NL into MCP/CLI calls) rather than the product itself having a native NL command interface, and write actions require human confirmation clicks rather than pure NL operation. missing for 10: evidence of a first-party conversational NL interface within Stripe's own Dashboard/product (not via third-party agent), and confirmation that full workflows can complete end-to-end via NL without manual approval steps.

   • [probe] “PROBE mcp-oauth (2026-09-14): keyless JSON-RPC initialize to https://mcp.stripe.com answers HTTP 401 {"error":"Unauthorized. See https://doc…”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “Docs, "Confirm actions by agents acting on your behalf": "To prevent agents from making mistakes, Stripe requires human confirmation before …”
   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”
   • [claimed-docs] “Official agent plugins bundle MCP + skills with auto-update: "Stripe's agent plugins give your AI agent access to Stripe developer tools, li…”
   • [claimed-docs] “Stripe's agent plugins give your AI agent access to Stripe developer tools, live account data, and the latest skills and technical guidance.”
   • [claimed-docs] “"Stripe Directory helps developers and AI agents find the best external providers for a task and follow the best supported path to provision…”
   • [claimed-docs] “The CLI landing page frames the command line as "Stripe's suite of agent-ready tooling", with "Agent setup: Automatically configure your age…”

   proof ↗⚑ flag

   Checkout.com✓fullprobed8/10

   Checkout.com ships an official MCP Server that lets users query payment statuses, issue refunds, void payments, and manage payment links via natural-language requests to an AI agent, plus guided onboarding through a 'Guide tool' — directly matching the story. This is corroborated by concrete endpoint/setup details (OAuth, sandbox/production URLs, `claude mcp add` command) rather than just marketing copy. Missing for 10: independent/hands-on user reports of the NL commands working (server is labeled Beta) and no evidence beyond vendor docs of reliability at scale.

   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”
   • [claimed-docs] “The Checkout.com Model Context Protocol (MCP) Server enables you to search through our knowledge base and manage payment operations directly…”
   • [probe] “official MCP server documented at https://www.checkout.com/docs/developer-resources/checkout-com-mcp-server”

   proof ↗⚑ flag

Api quality

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

   weight 2 · round to Stripe
   Stripe~partialprobed6/10

   Stripe's Workbench includes an in-Dashboard API Explorer/Shell that lets users visually explore API resources and execute real, runnable API calls (including against sandboxes), and Workbench's shell exposes CLI/API autocompletion — this is a genuine interactive, runnable API reference. However, evidence doesn't show a public-facing interactive API reference page (like a standalone docs.stripe.com/api with try-it-now runnable code snippets) accessible without Dashboard login, nor independent/community confirmation of the experience quality. missing for 10: evidence of a public runnable API reference outside the authenticated Dashboard, independent/hands-on user reviews of the API explorer experience, confirmation that all endpoints are runnable (not just Shell CLI-style commands).

   • [claimed-docs] “Workbench (Dashboard > Developers > Workbench) ships an interactive Shell and API Explorer: "Shell is a command line interface within Workbe…”
   • [claimed-docs] “Workbench provides powerful tools that help you build, test, and debug your Stripe integration.”
   • [probe] “PROBE docs-md: HTTP 200 at https://docs.stripe.com/payments.md # Payments Use Stripe to start accepting payments. ## Get started Integrat…”

   proof ↗⚑ flag

   Checkout.com—none0/10

   The evidence pack shows extensive documentation pages, an llms.txt feed, and an MCP server for AI agents, but no evidence of an interactive API reference with runnable/live code examples (e.g., a 'try it' console or embedded sandbox). The OpenAPI probe explicitly found all candidate spec paths returning 404, suggesting no machine-readable spec is exposed for such tooling.

   • [probe] “PROBE openapi: all candidate paths 404 (https://www.checkout.com/openapi.json, https://www.checkout.com/swagger.json, https://www.checkout.c…”
   • [probe] “PROBE docs-md: HTTP 404 at https://www.checkout.com/docs.md”
   • [claimed-docs] “Request a payment using a token, a network token, a payment instrument, or the full card details.”

   proof ↗⚑ flag

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

   weight 2 · round to Stripe
   Stripe✓fullprobed8/10

   Stripe publishes machine-readable OpenAPI spec files explicitly for generating SDKs/clients (stripe-docs-46, github.com/stripe/openapi), and docs.stripe.com also serves machine-readable .md versions of every doc page plus an llms.txt index (stripe-probe-1, stripe-probe-2), directly serving AI-native consumption needs. However, a direct openapi.json/swagger.json was not found at the docs.stripe.com domain itself (stripe-probe-3 shows 404s on guessed paths), meaning discovery requires knowing the separate GitHub repo rather than a single canonical documented download link. Missing for 10: a first-party documented URL/endpoint for the OpenAPI spec directly from docs.stripe.com (not just GitHub), and independent hands-on confirmation of spec completeness/versioning.

   • [claimed-docs] “Use these files to generate SDKs or client libraries that support Stripe's API.”
   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.stripe.com/llms.txt # Stripe Documentation When installing Stripe packages, always check the npm r…”
   • [probe] “PROBE docs-md: HTTP 200 at https://docs.stripe.com/payments.md # Payments Use Stripe to start accepting payments. ## Get started Integrat…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.stripe.com/openapi.json, https://docs.stripe.com/swagger.json, https://docs.stripe.com/…”

   proof ↗⚑ flag

   Checkout.com—none0/10

   The evidence pack shows an llms.txt (LLM-friendly documentation index) but explicit probes for OpenAPI/Swagger specs at common paths (openapi.json, swagger.json, .well-known/openapi.json) all returned 404, and no other citation references a downloadable OpenAPI or equivalent machine-readable API spec.

   • [probe] “PROBE openapi: all candidate paths 404 (https://www.checkout.com/openapi.json, https://www.checkout.com/swagger.json, https://www.checkout.c…”
   • [probe] “PROBE llms.txt: HTTP 200 at https://www.checkout.com/llms.txt # Checkout.com: High-performance global payments platform for enterprise and s…”
   • [probe] “PROBE docs-md: HTTP 404 at https://www.checkout.com/docs.md”

   proof ↗⚑ flag

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

   weight 1 · round to Stripe
   Stripe✓fullclaimed9/10

   Stripe documents a dedicated sandbox environment isolated from live/production data (stripe-docs-32, stripe-docs-43, stripe-docs-57), test cards/values that simulate transactions without moving real funds (stripe-docs-10, stripe-docs-65), and anonymous sandbox provisioning explicitly for coding agents via CLI (stripe-docs-16), plus MCP/Workbench access scoped per sandbox vs live mode (stripe-supp-mcp-sessions-governance, stripe-supp-workbench-shell). Missing for 10: independent hands-on confirmation from a third party that agent-driven sandbox testing works end-to-end without friction.

   • [claimed-docs] “A sandbox is an isolated test environment. You can use your sandbox to test Stripe functionality in your account, and experiment with new fe…”
   • [claimed-docs] “You can use your sandbox to test Stripe functionality in your account, and experiment with new features without affecting your live integrat…”
   • [claimed-docs] “Your team can test in separate sandboxes to make sure that data and actions are completely isolated from other sandboxes.”
   • [claimed-docs] “Test your integration in a sandbox ... by simulating transactions with test values—these transactions don't move funds.”
   • [claimed-docs] “Test cards act as "fake" credit cards, and allow you to simulate the following scenarios: - Successful payments... - Card errors due to decl…”
   • [claimed-docs] “Coding agents should install the Stripe CLI (npm i -g @stripe/cli) and run the command stripe sandbox create --help to provision an anonymou…”
   • [claimed-docs] “MCP access is governed like any credential: OAuth per the MCP spec ("OAuth lets you grant and revoke access without sharing an API key with …”
   • [claimed-docs] “Workbench (Dashboard > Developers > Workbench) ships an interactive Shell and API Explorer: "Shell is a command line interface within Workbe…”
   • [claimed-docs] “The CLI landing page frames the command line as "Stripe's suite of agent-ready tooling", with "Agent setup: Automatically configure your age…”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Checkout.com documents a dedicated sandbox/test-account environment (test cards, disputes testing, fraud/payouts testing, and a public 'get test account' signup) separate from production, and the MCP server itself is explicitly split into a sandbox endpoint (checkout.mcp.sbox.cko.tech) versus production (mcp.checkout.com), letting AI-native integrations test agentic operations without touching live data. Missing for 10: independent/hands-on confirmation that the sandbox fully mirrors production behavior and more detail on data isolation guarantees.

   • [claimed-docs] “A dedicated test-account environment with per-scenario testing docs: payments testing, test cards, disputes testing, fraud-detection testing…”
   • [claimed-docs] “Test cards”
   • [claimed-docs] “Disputes testing”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”

   proof ↗⚑ flag

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

   weight 2 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe documents a clear API versioning policy: monthly releases contain only backward-compatible changes, and users can safely upgrade without breaking existing code, with explicit versioning docs. This directly matches the story's requirement of versioned APIs with a documented deprecation/compatibility policy. Missing for 10: no explicit deprecation timeline/sunset schedule detail or independent hands-on verification of long-term version support beyond the docs' own claims.

   • [claimed-docs] “You can safely upgrade to a new monthly release without breaking any existing code.”
   • [claimed-docs] “Each monthly release includes only backward-compatible changes... You can safely upgrade to a new monthly release without breaking any exist…”

   proof ↗⚑ flag

   Checkout.com—none0/10

   The evidence pack covers idempotency, rate limits, auth, webhooks, and testing, but contains no mention of API version headers, a changelog, or a documented deprecation/sunset policy; probes for an OpenAPI spec even returned 404s, suggesting no discoverable versioning artifact.

   ⚑ flag

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

   weight 2 · round to Stripe
   Stripe~partialclaimed6/10

   Stripe supports some bulk-style operations relevant to AI agents — bulk product catalog import via the v2 Product Catalog Import API (CSV up to 4GB, upsert/replace modes), List CheckoutSessions for bulk fulfillment, and MCP/analytics tools that can query/report across many objects (stripe_analytics, stripe_report) — but there is no documented generic bulk-write or batch-update API (e.g., bulk refunds, bulk subscription updates, bulk customer edits) that an AI agent could invoke across many arbitrary items at once; most single-object write tools (stripe_api_write, Refunds API, etc.) operate one record at a time with idempotency keys rather than true batch semantics. missing for 10: a documented bulk/batch API for core objects (customers, charges, subscriptions, refunds) beyond catalog import, evidence of MCP or CLI tools performing multi-item operations in one call, and independent/hands-on confirmation of bulk operation reliability.

   • [claimed-docs] “Agentic Commerce Suite ingests full product catalogs as bulk CSV imports via the v2 Product Catalog Import API: POST /v2/commerce/product_ca…”
   • [claimed-docs] “Orders placed through AI chat agents flow through standard Stripe rails: "Stripe sends checkout.session.completed after the agent completes …”
   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”

   proof ↗⚑ flag

   Checkout.com—none0/10

   The evidence pack covers single-resource operations (payments, refunds, payment links, webhooks) and idempotent retries for individual requests, but there is no documentation of a batch/bulk API, bulk refund/payout endpoints, or MCP tooling that operates across many items at once. Reconciliation and reporting are read-only aggregate views, not bulk write operations.

   ⚑ flag

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

   weight 3 · round to Stripe
   Stripe~partialclaimed6/10

   Stripe provides the event backbone needed for rule-based automation — webhooks/event destinations that fire on account events (stripe-docs-8, stripe-docs-53, stripe-supp-event-destinations) — plus some built-in automated actions like Smart Retries and Radar's real-time risk rules (stripe-docs-14, stripe-docs-22). A third-party Dashboard app (Kyber) explicitly offers a no-code rule builder that 'assigns tasks and automates common workflows based on Stripe events' (e.g., Slack alert on dispute), which matches the story closely (stripe-intdir-6). However, this rule-definition UI is a third-party app, not a native first-party Stripe rules engine, and there's no evidence of a built-in general-purpose 'if event X then action Y' configurator inside Stripe itself. Missing for 10: a native first-party no-code/low-code rule engine for arbitrary event-triggered actions, and independent hands-on confirmation of such rule-based automation in practice.

   • [claimed-docs] “Receiving webhook events helps you respond to asynchronous events, such as when a customer's bank confirms a payment, a customer disputes a …”
   • [claimed-docs] “You can create an HTTPS webhook endpoint to receive events.”
   • [claimed-docs] “After you register a webhook endpoint, Stripe pushes real-time data to it when events happen in your Stripe account.”
   • [claimed-docs] “Use Radar to evaluate transactions, accounts, and customers in real time, using AI algorithms to assess the risk of fraud.”
   • [claimed-docs] “Smart Retries: Automatically retry failed payments to prevent involuntary churn due to temporary issues.”
   • [claimed-docs] “The no-code builder Kyber lets users assign tasks and automate common workflows based on Stripe events. Users can quickly configure automati…”
   • [claimed-docs] “Event destinations deliver real-time events "across multiple destination types, including webhook endpoints, Amazon EventBridge, and Azure E…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed4/10

   Checkout.com supports webhooks that notify external systems on payment lifecycle events, and documented 'risk rules and lists' that automatically trigger fraud-prevention actions based on transaction conditions — both are forms of event-triggered automation. However, there is no evidence of a general-purpose, user-defined rules/automation engine (e.g., 'if X event then Y action' workflow builder) spanning arbitrary events beyond fraud, so it only partially satisfies the story. Missing for 10: a native no-code/low-code rules engine for arbitrary event-triggered actions, documented conditional logic builder, and evidence of automation beyond fraud rules and webhook delivery.

   • [claimed-docs] “Create risk rules and lists”
   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “Receive webhooks”
   • [claimed-docs] “Retrieve fraud reports”

   proof ↗⚑ flag

3. ai-native userSchedule recurring jobs or workflows

   weight 2 · round to Stripe
   Stripe~partialclaimed6/10

   Stripe Billing natively automates recurring payment jobs (subscriptions, invoicing, trials, renewals, smart retries) and this can be driven via API/MCP by an AI agent, plus Stripe Projects/CLI/Workbench let agents script and schedule integration tasks. However there is no evidence of a general-purpose 'schedule any workflow/job' primitive (e.g. cron-like task scheduler for arbitrary agent workflows) beyond payment/billing-specific recurrence. missing for 10: a generic workflow/job scheduler beyond billing recurrence, evidence of agents autonomously scheduling multi-step workflows (not just recurring payments), independent hands-on confirmation of this automation depth.

   • [claimed-docs] “It automates recurring payments, creates custom pricing plans, and handles billing periods, such as trials and renewals.”
   • [claimed-docs] “Smart Retries: Automatically retry failed payments to prevent involuntary churn due to temporary issues.”
   • [claimed-docs] “Use Stripe Billing to manage subscriptions and invoicing. It automates recurring payments, creates custom pricing plans, and handles billing…”
   • [claimed-docs] “Stripe Projects "provisions and manages third-party services (such as hosting, databases, auth, AI, and observability) from the terminal. Ru…”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “You can use an AI agent to analyze aggregated data across your Stripe account ... query metrics such as monthly recurring revenue (MRR)”

   proof ↗⚑ flag

   Checkout.com—none0/10

   The evidence pack documents payment processing, webhooks, MCP-driven operations, and reconciliation reports, but nowhere describes recurring/scheduled job or workflow automation (e.g., subscription billing, cron-like scheduling, recurring payment automation) that an AI-native user could set up. No mention of scheduling, recurring billing, or workflow orchestration exists in the pack.

   ⚑ flag

4. ai-native userVersion, review, and roll back my automations

   weight 1 · round drawn
   Stripe—none0/10

   Stripe's automation features (Billing recurring rules, Smart Retries, Radar rules, webhooks, Kyber no-code flows) have no documented version-history, review/diff, or rollback mechanism in the evidence. API versioning (stripe-docs-45/59) covers backward-compatible monthly API releases, not versioning of a user's automation configurations, so it doesn't satisfy this story.

   • [claimed-docs] “You can safely upgrade to a new monthly release without breaking any existing code.”
   • [claimed-docs] “Each monthly release includes only backward-compatible changes... You can safely upgrade to a new monthly release without breaking any exist…”
   • [claimed-docs] “Smart Retries: Automatically retry failed payments to prevent involuntary churn due to temporary issues.”
   • [claimed-docs] “Use Stripe Billing to manage subscriptions and invoicing. It automates recurring payments, creates custom pricing plans, and handles billing…”
   • [claimed-docs] “The no-code builder Kyber lets users assign tasks and automate common workflows based on Stripe events. Users can quickly configure automati…”

   proof ↗⚑ flag

   Checkout.com—none0/10

   Checkout.com is a payments platform; the evidence covers webhooks, idempotency, MCP server, and flows/plugins, but nothing documents versioning, review workflows, or rollback of 'automations' (e.g. Flow configs, webhook rules, risk rules) as artifacts with history/diff/rollback capability.

   ⚑ flag

Ai billing ops

1. ai-native userHave an agent monitor failed payments via API and webhooks and drive retries and dunning outreach

   weight 3 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe documents the full loop an agent needs: webhooks/event destinations for detecting failed payments and dispute events (stripe-docs-8, stripe-docs-53, stripe-supp-event-destinations), Smart Retries and no-code revenue recovery/dunning automation with recovery analytics (stripe-docs-14, stripe-docs-15, stripe-docs-56), and API/MCP write access (subscriptions, invoices, disputes) with restricted-key scoping and human-confirmation guardrails so an agent can drive retries and outreach programmatically (stripe-supp-mcp-tools, stripe-supp-mcp-write-confirmation, stripe-supp-restricted-keys-agents, stripe-docs-41). Idempotency support further protects automated retry logic (stripe-docs-9). missing for 10: no first-party or hands-on example of an agent specifically orchestrating dunning outreach (e.g., emails/SMS) end-to-end, and no independent/community confirmation of this exact agent workflow in production.

   • [claimed-docs] “Receiving webhook events helps you respond to asynchronous events, such as when a customer's bank confirms a payment, a customer disputes a …”
   • [claimed-docs] “You can create an HTTPS webhook endpoint to receive events.”
   • [claimed-docs] “Smart Retries: Automatically retry failed payments to prevent involuntary churn due to temporary issues.”
   • [claimed-docs] “Recovery analytics: Analyze subscription payment failure rates, recovery rates, and recent failed payments for top customers.”
   • [claimed-docs] “Prevent lost revenue and reduce churn with Stripe’s revenue recovery features... None of the features require you to write code.”
   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “Docs, "Confirm actions by agents acting on your behalf": "To prevent agents from making mistakes, Stripe requires human confirmation before …”
   • [claimed-docs] “Restricted API keys are Stripe's documented least-privilege credential for agents: "When you create a RAK in the Stripe Dashboard, you selec…”
   • [claimed-docs] “Event destinations deliver real-time events "across multiple destination types, including webhook endpoints, Amazon EventBridge, and Azure E…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed4/10

   Checkout.com provides the raw building blocks an agent could use — webhooks for payment lifecycle events (including failures), idempotent retry-safe API endpoints, and an MCP server that can query payment statuses and manage payment operations from an AI agent — but there is no documented dunning/retry-campaign feature or subscription-billing failed-payment workflow (no mention of automatic retry scheduling, dunning emails, or a 'monitor failed payments and drive outreach' capability). missing for 10: dedicated dunning/retry-campaign automation, evidence of subscription/recurring billing failure handling, and any documented agent workflow specifically for failed-payment monitoring/outreach.

   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “Idempotent retries are first-class: "The following endpoints from our API support idempotent requests, so you can safely retry them without …”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”
   • [claimed-docs] “Receive webhooks”
   • [claimed-docs] “Resend webhooks”

   proof ↗⚑ flag

2. ai-native userHave an agent provision a new product, price, and payment link via the API in a single session

   weight 3 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe's MCP server documents stripe_api_write with create/update support for products, prices, and payment links, plus stripe_implementation_planner to guide multi-step builds, enabling an agent to provision a product, price, and payment link in one session via a single authenticated MCP/API connection; the CLI/sandbox provisioning (stripe sandbox create) also lets agents get working keys without manual account setup. Missing for 10: a hands-on end-to-end transcript/independent report of an agent actually completing this exact three-step chain in one session, and clarity on whether human-confirmation gating (required for refunds/payments) also intercepts routine create-product/price/payment-link calls.

   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “Coding agents should install the Stripe CLI (npm i -g @stripe/cli) and run the command stripe sandbox create --help to provision an anonymou…”
   • [claimed-docs] “Docs, "Confirm actions by agents acting on your behalf": "To prevent agents from making mistakes, Stripe requires human confirmation before …”
   • [claimed-docs] “Create a payment link and share it on social media, in emails, or on your website.”
   • [claimed-docs] “The CLI landing page frames the command line as "Stripe's suite of agent-ready tooling", with "Agent setup: Automatically configure your age…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed4/10

   Checkout.com documents API-driven Payment Link creation (docs-19/56) and an MCP server that can be used from an AI-assisted IDE to 'create and manage payment links' (cko-supp-mcp-server), so an agent could plausibly create a payment link in one session via API/MCP. However, there is no evidence of a discrete 'product' or 'price' catalog object/API (unlike a full product-catalog model) — Checkout.com's docs only reference amount/currency fields on payment links, not reusable Product/Price entities. missing for 10: a documented Products API and Prices API (or catalog object model) that an agent could provision alongside a payment link, and any hands-on/independent confirmation of an agent completing product+price+link creation end-to-end in one session.

   • [claimed-docs] “Create a Payment Link”
   • [claimed-docs] “Create a Payment Link - Docs”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”

   proof ↗⚑ flag

Subscriptions

1. finance-ops userAutomatically retry failed subscription payments with smart dunning and recovery emails

   weight 2 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe Billing's revenue-recovery docs explicitly document Smart Retries for automatically retrying failed subscription payments plus recovery analytics, and this is a no-code feature within Billing. Missing for 10: explicit documentation of automated dunning/recovery email templates and independent hands-on validation of recovery email effectiveness.

   • [claimed-docs] “Smart Retries: Automatically retry failed payments to prevent involuntary churn due to temporary issues.”
   • [claimed-docs] “Recovery analytics: Analyze subscription payment failure rates, recovery rates, and recent failed payments for top customers.”
   • [claimed-docs] “None of the features require you to write code, so you can start recovering revenue today.”
   • [claimed-docs] “Prevent lost revenue and reduce churn with Stripe’s revenue recovery features... None of the features require you to write code.”
   • [claimed-docs] “Prevent lost revenue and reduce churn with Stripe's revenue recovery features. ... None of the features require you to write code”
   • [claimed-docs] “It automates recurring payments, creates custom pricing plans, and handles billing periods, such as trials and renewals.”
   • [claimed-docs] “Use Stripe Billing to manage subscriptions and invoicing. It automates recurring payments, creates custom pricing plans, and handles billing…”

   proof ↗⚑ flag

   Checkout.com—none0/10

   The evidence pack shows Checkout.com as a payments/acquiring platform with API idempotency (for network retries, not subscription dunning), webhooks, refunds, and platforms/marketplace features, but there is no documented subscription billing product, dunning logic, automatic retry scheduling for failed recurring payments, or recovery email capability. This axis is plausible for a payments platform, but no evidence supports it being delivered.

   • [claimed-docs] “Idempotent retries are first-class: "The following endpoints from our API support idempotent requests, so you can safely retry them without …”
   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “The Platforms solution enables marketplaces, payment facilitators (Payfacs), and other platform-based businesses to onboard sub-entities and…”

   proof ↗⚑ flag

2. developerSet up recurring subscription billing with plans, trials, upgrades, and proration

   weight 3 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe Billing explicitly automates recurring payments, custom pricing plans, trials, renewals, and subscription upgrades, plus a customer portal for self-service management, revenue recovery (smart retries), and subscription migration docs. Documentation confirms core subscription lifecycle features via first-party docs, though proration is only implied via 'upgrades' rather than explicitly named, and no independent/hands-on corroboration of proration mechanics is present. Missing for 10: explicit proration documentation/examples, independent developer corroboration of subscription setup experience.

   • [claimed-docs] “It automates recurring payments, creates custom pricing plans, and handles billing periods, such as trials and renewals.”
   • [claimed-docs] “Use Stripe Billing to manage subscriptions and invoicing. It automates recurring payments, creates custom pricing plans, and handles billing…”
   • [claimed-docs] “Use Stripe Billing to manage subscriptions and invoicing. It automates recurring payments, creates custom pricing plans, and handles billing…”
   • [claimed-docs] “Integrate the customer portal: The simplest way to build a subscription and billing management Dashboard.”
   • [claimed-docs] “Migrate your subscriptions: Learn how to migrate subscriptions to Stripe.”
   • [claimed-docs] “Smart Retries: Automatically retry failed payments to prevent involuntary churn due to temporary issues.”
   • [claimed-docs] “Recovery analytics: Analyze subscription payment failure rates, recovery rates, and recent failed payments for top customers.”
   • [claimed-docs] “None of the features require you to write code, so you can start recovering revenue today.”

   proof ↗⚑ flag

   Checkout.com—none0/10

   Checkout.com is a payments/acquiring platform (payment processing, links, hosted pages, platforms/marketplaces, issuing) but the evidence pack contains no mention of subscription/recurring billing objects such as plans, trial periods, proration, or plan upgrades/downgrades — these are billing-engine concepts (like Stripe Billing/Chargebee) not documented here.

   ⚑ flag

Card acceptance

1. founderAccept an online card payment within a day of signing up

   weight 3 · round to Checkout.com
   Stripe!disputedcontradicted6/10

   Stripe's docs strongly support rapid onboarding: no-code Payment Links, hosted Checkout pages, and a live testimonial confirms setup and recurring payments working 'in less than an hour' (stripe-docs-4, stripe-docs-2, stripe-comm-9). However, independent community reports concretely contradict universal same-day acceptance — users report Stripe denying account creation entirely or 'nuking' an existing business's payment capability without quick support resolution (stripe-comm-1, stripe-comm-4, stripe-comm-5), showing that for some founders (especially certain geographies/business types) the promised fast, frictionless activation does not materialize. missing for 10: evidence resolving how common these denials/delays are, and first-party docs addressing account approval timelines or geographic/business restrictions.

   • [claimed-docs] “Create a payment link and share it on social media, in emails, or on your website.”
   • [claimed-docs] “Customers enter their payment details in a fully-featured payment page, either embedded on your site or via a redirect to a Stripe-hosted pa…”
   • [claimed-docs] “Create a payments page with prebuilt UIs using the Checkout Sessions API.”
   • [community] “Stripe is a game changer. I've been using it for a few months and honestly its the best API I've ever used. The documentation is clear and c…”
   • [community] “Funny, I have a Delaware LLC company registered in USA (it's an offshore company) but my bank account is outside the US (completely legal), …”
   • [community] “Stripe has decided to nuke my entire business... So currently, my product doesn't work for 35% of my customers. Cue torrent of pissed off cu…”
   • [community] “I wanted to enable issuing on my account to test card issuing, but there was a bug where the UI was looping. Asked support, got 2 weeks of '…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed6/10

   Checkout.com explicitly markets a Hosted Payment Page you can get 'fully branded... in less than a day' and offers self-serve test accounts, payment links, and low-code Flow for fast integration, suggesting a founder could technically wire up a checkout quickly. However, the pricing model is explicitly sales-led with 'no self-serve rate card' and merchants must 'get in touch with our team' to create a plan, implying an onboarding/underwriting step before going live with real payments that isn't addressed in the evidence. Missing for 10: evidence of self-serve signup-to-live-payment flow without sales contact, KYC/merchant approval timelines, and confirmation that a brand-new account can process a live card charge within 24 hours.

   • [claimed-docs] “Get a fully branded checkout page in less than a day. No heavy lifting or confusing compliance. Just one integration, and you're ready to ta…”
   • [claimed-docs] “A dedicated test-account environment with per-scenario testing docs: payments testing, test cards, disputes testing, fraud-detection testing…”
   • [claimed-docs] “Pricing is sales-led and unpublished: a tailored "Fully Flat-Rate" plan ("We price based on your business profile and risk category, nothing…”
   • [claimed-docs] “Create a Payment Link”
   • [claimed-docs] “Flow is a clever bit of code that helps you boost conversions, stay compliant, and enter new markets fast. Accept payments using customizabl…”

   proof ↗⚑ flag

Checkout surfaces

1. developerEmbed customizable payment components directly in my own site or app instead of redirecting to a hosted page

   weight 3 · round to Stripe
   Stripe✓fullclaimed9/10

   Stripe Elements and Checkout Sessions API let developers embed a fully customizable payment page directly in their own site (as opposed to only redirecting to a hosted page), with support for 125+ payment methods, subscriptions, and prebuilt/custom UI options documented extensively. Missing for 10: no independent/hands-on developer report specifically praising the embedded (non-redirect) Elements experience beyond docs.

   • [claimed-docs] “Build a fully customized payment page using elements”
   • [claimed-docs] “Customers enter their payment details in a fully-featured payment page, either embedded on your site or via a redirect to a Stripe-hosted pa…”
   • [claimed-docs] “Elements: Build a fully customized payment page using elements”
   • [claimed-docs] “Create a payments page with prebuilt UIs using the Checkout Sessions API.”
   • [claimed-docs] “Accept one-time and subscription payments from more than 125 local payment methods.”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed7/10

   Checkout.com's Flow product is explicitly documented as embeddable, customizable payment components in a single integration rather than a redirect ('Accept payments using customizable components – all through one simple integration'), alongside Mobile SDKs for native embedded checkout. missing for 10: hands-on/independent verification of the embedded component customization (theming depth, framework support), and a code sample showing actual component embedding rather than marketing copy.

   • [claimed-docs] “Flow is a clever bit of code that helps you boost conversions, stay compliant, and enter new markets fast. Accept payments using customizabl…”
   • [claimed-docs] “Get started with low-code Flow”
   • [claimed-docs] “Take charge of your mobile checkout with Flow for Mobile. Available via iOS and Android Mobile SDKs.”
   • [claimed-docs] “Mobile SDK: Build native mobile checkout experiences for iOS and Android.”

   proof ↗⚑ flag

2. developerRedirect customers to a vendor-hosted, conversion-optimized checkout page without building my own payment UI

   weight 3 · round to Stripe
   Stripe✓fullclaimed9/10

   Docs explicitly describe Stripe Checkout as a Stripe-hosted redirect page ('via a redirect to a Stripe-hosted page') with prebuilt UI via Checkout Sessions API, plus Payment Links for a fully no-code hosted checkout URL — directly matching the story. Multiple docs reinforce it supports many payment methods, currencies, tax collection, and no-code refunds on these hosted pages. missing for 10: no independent/hands-on corroboration of the hosted page's conversion-optimization claims beyond Stripe's own docs.

   • [claimed-docs] “Customers enter their payment details in a fully-featured payment page, either embedded on your site or via a redirect to a Stripe-hosted pa…”
   • [claimed-docs] “Create a payments page with prebuilt UIs using the Checkout Sessions API.”
   • [claimed-docs] “Create a payment link and share it on social media, in emails, or on your website.”
   • [claimed-docs] “Let your customers use a link to pay in their local language and currency with their preferred payment method.”
   • [claimed-docs] “Collect tax on Payment Links: Learn how to calculate and collect tax on a payment page without writing any code.”
   • [claimed-docs] “Accept one-time and subscription payments from more than 125 local payment methods.”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Checkout.com documents a dedicated Hosted Payments Page product ('Get a fully branded checkout page in less than a day... one integration') with low-code deployment and customization docs, plus a related Payment Links option for no-code hosted checkout flows. This directly matches the story of redirecting to a vendor-hosted, conversion-optimized checkout page without building custom payment UI. Missing for 10: independent/third-party conversion benchmarks or hands-on merchant testimonials validating the 'conversion-optimized' claim beyond vendor copy.

   • [claimed-docs] “Get a fully branded checkout page in less than a day. No heavy lifting or confusing compliance. Just one integration, and you're ready to ta…”
   • [claimed-docs] “Customize your Hosted Payments Page”
   • [claimed-docs] “Hosted Payment Page: Deploy a pre-built, low-code payment page integration.”
   • [claimed-docs] “Create a Payment Link”
   • [claimed-docs] “Send secure Payment Links over email, SMS, social, and more – without any need for coding – and open up more revenue opportunities, no matte…”
   • [claimed-docs] “Flow is a clever bit of code that helps you boost conversions, stay compliant, and enter new markets fast. Accept payments using customizabl…”

   proof ↗⚑ flag

No code selling

1. founderCreate a shareable payment link from the dashboard with no code and start selling immediately

   weight 2 · round to Stripe
   Stripe✓fullclaimed9/10

   Stripe Payment Links is explicitly documented as a no-code way to create a shareable payment link from the dashboard, supporting local currencies/payment methods, no-code tax collection, and no-code refunds — matching the founder story closely. Missing for 10: independent hands-on account of actually creating a payment link end-to-end (community evidence is general Stripe sentiment, not specific to Payment Links).

   • [claimed-docs] “Create a payment link and share it on social media, in emails, or on your website.”
   • [claimed-docs] “No-code refunds Yes”
   • [claimed-docs] “Let your customers use a link to pay in their local language and currency with their preferred payment method.”
   • [claimed-docs] “Collect tax on Payment Links: Learn how to calculate and collect tax on a payment page without writing any code.”

   proof ↗⚑ flag

   Checkout.com~partialclaimed6/10

   Checkout.com explicitly documents no-code Payment Links that can be sent via email/SMS/social to accept payments without coding (checkout-com-docs-5, -19, -41, -56), which directly matches the story. However, the account/pricing evidence indicates a sales-led, KYC-gated onboarding model rather than instant self-serve signup ('get in touch with our team' — cko-supp-pricing), undercutting the 'start selling immediately' claim for a new founder. Missing for 10: evidence of an instant self-serve dashboard signup (no sales call) and independent/hands-on confirmation that a first-time user can generate a link within minutes.

   • [claimed-docs] “Payment Links: Accept payments on any channel using shareable payment links.”
   • [claimed-docs] “Create a Payment Link”
   • [claimed-docs] “Send secure Payment Links over email, SMS, social, and more – without any need for coding – and open up more revenue opportunities, no matte…”
   • [claimed-docs] “Create a Payment Link - Docs”
   • [claimed-docs] “Pricing is sales-led and unpublished: a tailored "Fully Flat-Rate" plan ("We price based on your business profile and risk category, nothing…”

   proof ↗⚑ flag

Payment methods

1. founderOffer digital wallets and local payment methods (Apple Pay, Google Pay, bank redirects, BNPL) from the same integration

   weight 2 · round to Checkout.com
   Stripe~partialclaimed7/10

   Stripe Checkout/Elements documentation confirms 125+ local payment methods and localized payment links, which covers digital wallets, bank redirects, and local methods from one integration; however, the evidence never explicitly names Apple Pay, Google Pay, or specific BNPL providers, and there's no independent/hands-on confirmation of the exact method mix. missing for 10: explicit mention of Apple Pay/Google Pay support, named BNPL providers (e.g., Klarna/Afterpay), and independent verification of the full payment-method list.

   • [claimed-docs] “Accept one-time and subscription payments from more than 125 local payment methods.”
   • [claimed-docs] “Build a fully customized payment page using elements”
   • [claimed-docs] “Customers enter their payment details in a fully-featured payment page, either embedded on your site or via a redirect to a Stripe-hosted pa…”
   • [claimed-docs] “Let your customers use a link to pay in their local language and currency with their preferred payment method.”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Checkout.com documents a unified payments API and Flow/Hosted Payment Page integrations that give access to digital wallets, local payment methods, and global card schemes via a single integration (checkout-com-docs-45, checkout-com-docs-2, checkout-com-docs-38, checkout-com-docs-43), explicitly covering local currencies and regional payment strategies. Missing for 10: no explicit line-item confirmation of BNPL providers or bank redirect method names, and no independent/hands-on corroboration of wallet coverage breadth.

   • [claimed-docs] “Get instant access to all the local payment methods, digital wallets, and global card schemes you need for smarter regional strategies and h…”
   • [claimed-docs] “Unified Payments API: Connect to all payment methods through a single API integration.”
   • [claimed-docs] “Flow is a clever bit of code that helps you boost conversions, stay compliant, and enter new markets fast. Accept payments using customizabl…”
   • [claimed-docs] “Take complete control of the entire payment lifecycle from a single API. Upgrade, expand, and adapt your payment experiences with a flexible…”
   • [claimed-docs] “Accept payments like a local business globally. Get paid in local currencies, settle funds fast, and gain valuable insights.”
   • [claimed-docs] “Process worldwide in 150+ currencies. Plus domestic coverage in 45+ countries and counting.”

   proof ↗⚑ flag

Pci compliance

1. developerKeep my systems out of most of PCI DSS scope because card data is tokenized and never touches my servers

   weight 2 · round to Stripe
   Stripe✓fullclaimed9/10

   Stripe's docs explicitly state that low-risk integrations (Checkout, Elements, hosted payment pages) collect and transmit payment data directly to Stripe without it passing through the developer's servers, explicitly reducing PCI obligations/scope, and these are backed by concrete hosted/embeddable products (Checkout, Payment Links, Elements). missing for 10: independent third-party PCI audit or compliance-report corroboration beyond Stripe's own docs.

   • [claimed-docs] “You can instead use one of our low risk payment integrations to securely collect and transmit payment information directly to Stripe without…”
   • [claimed-docs] “you can instead use one of our low risk payment integrations to securely collect and transmit payment information directly to Stripe without…”
   • [claimed-docs] “use one of our low risk payment integrations to securely collect and transmit payment information directly to Stripe without it passing thro…”
   • [claimed-docs] “Build a fully customized payment page using elements”
   • [claimed-docs] “Customers enter their payment details in a fully-featured payment page, either embedded on your site or via a redirect to a Stripe-hosted pa…”
   • [claimed-docs] “Create a payment link and share it on social media, in emails, or on your website.”

   proof ↗⚑ flag

   Checkout.com~partialclaimed6/10

   Checkout.com offers hosted payment pages, Flow, mobile SDKs, and payment links that are explicitly marketed as reducing compliance burden ('no confusing compliance'), and also supports tokens/network tokens/payment instruments as an alternative to raw card data, all of which are standard mechanisms for keeping card data off merchant servers and reducing PCI scope. However, no evidence explicitly names PCI DSS, SAQ levels, or describes tokenization architecture/vaulting in detail, and the API also documents accepting 'full card details' directly, meaning scope reduction depends on which integration path a developer chooses. Missing for 10: explicit PCI DSS compliance/SAQ-A scope-reduction documentation, dedicated tokenization/vaulting API docs, and independent confirmation of scope-reduction claims.

   • [claimed-docs] “Request a payment using a token, a network token, a payment instrument, or the full card details.”
   • [claimed-docs] “Flow is a clever bit of code that helps you boost conversions, stay compliant, and enter new markets fast. Accept payments using customizabl…”
   • [claimed-docs] “Get a fully branded checkout page in less than a day. No heavy lifting or confusing compliance. Just one integration, and you're ready to ta…”
   • [claimed-docs] “Take charge of your mobile checkout with Flow for Mobile. Available via iOS and Android Mobile SDKs.”
   • [claimed-docs] “Send secure Payment Links over email, SMS, social, and more – without any need for coding – and open up more revenue opportunities, no matte…”
   • [claimed-docs] “Customize your Hosted Payments Page”

   proof ↗⚑ flag

Integration reliability

1. developerSafely retry any API request using idempotency keys so network failures never create duplicate charges

   weight 2 · round drawn
   Stripe✓fullclaimed9/10

   Stripe's API docs explicitly document idempotency keys for safely retrying requests after connection errors without creating duplicate objects/charges, directly matching the story, and this is a well-known, widely used first-party API feature. Missing for 10: independent hands-on developer corroboration specifically testing idempotency behavior (community evidence in the pack is unrelated to this feature).

   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”
   • [claimed-docs] “use an idempotency key... you can safely repeat the request without risk of creating a second object or performing the update twice”
   • [claimed-docs] “use an idempotency key. Then, if a connection error occurs, you can safely repeat the request without risk of creating a second object or pe…”
   • [claimed-docs] “use an idempotency key. Then, if a connection error occurs, you can safely repeat the request without risk of creating a second object”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed9/10

   Checkout.com's docs explicitly document idempotency support via the Cko-Idempotency-Key header across payments, authorizations, captures, refunds, cancellations, voids, and transfers endpoints, caching results of 2xx responses to safely retry without duplicate charges, with a configurable idempotency window. Missing for 10: independent/hands-on developer corroboration of real-world retry behavior beyond vendor docs.

   • [claimed-docs] “Idempotent retries are first-class: "The following endpoints from our API support idempotent requests, so you can safely retry them without …”
   • [claimed-docs] “Occasionally, you may need to retry a request due to a disrupted API call. For example, if you encounter a network error or timeout.”
   • [claimed-docs] “Idempotency - Docs”

   proof ↗⚑ flag

2. developerRely on signed webhook deliveries with automatic retries, event logs, and replay to keep my system consistent with payment state

   weight 3 · round to Stripe
   Stripe✓fullclaimed8/10

   Docs cover signed/HTTPS webhook endpoints, real-time event pushes, idempotency for safe retries, dispute/event handling via API+webhooks, and the newer event destinations feature with versioned snapshot/thin events; sandbox testing supports simulating webhook flows. However, explicit documentation of automatic retry schedules for failed webhook deliveries and a dedicated 'replay' UI/feature isn't directly cited (event destinations mention versioning but not an explicit replay tool), and there's no independent/hands-on corroboration of retry reliability. missing for 10: explicit docs on automatic webhook retry backoff schedule, a dedicated event replay feature/UI, independent developer corroboration of retry/replay reliability.

   • [claimed-docs] “Receiving webhook events helps you respond to asynchronous events, such as when a customer's bank confirms a payment, a customer disputes a …”
   • [claimed-docs] “Stripe pushes real-time data to it when events happen in your Stripe account.”
   • [claimed-docs] “You can create an HTTPS webhook endpoint to receive events.”
   • [claimed-docs] “After you register a webhook endpoint, Stripe pushes real-time data to it when events happen in your Stripe account.”
   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”
   • [claimed-docs] “use an idempotency key... you can safely repeat the request without risk of creating a second object or performing the update twice”
   • [claimed-docs] “use an idempotency key. Then, if a connection error occurs, you can safely repeat the request without risk of creating a second object or pe…”
   • [claimed-docs] “use an idempotency key. Then, if a connection error occurs, you can safely repeat the request without risk of creating a second object”
   • [claimed-docs] “With the API, you can upload evidence, respond to disputes, and receive dispute events using webhooks.”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”
   • [claimed-docs] “Event destinations deliver real-time events "across multiple destination types, including webhook endpoints, Amazon EventBridge, and Azure E…”
   • [claimed-docs] “Workbench provides powerful tools that help you build, test, and debug your Stripe integration.”

   proof ↗⚑ flag

   Checkout.com~partialclaimed6/10

   Checkout.com documents HMAC-signed webhooks, dashboard/API-based webhook configuration, and a dedicated 'Resend webhooks' capability for replay, satisfying the signing and replay parts of the story. However, evidence does not confirm automatic system-driven retries of failed webhook deliveries (only manual resend) or a dedicated event log/audit trail view for webhook history. Missing for 10: documented automatic retry policy for undelivered webhooks, explicit event log/audit UI, independent confirmation of retry behavior.

   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “Receive webhooks”
   • [claimed-docs] “Resend webhooks”
   • [claimed-docs] “Manage webhooks in the Dashboard”

   proof ↗⚑ flag

Onboarding dx

1. developerIntegrate a working payment flow from the docs alone in under a day, with copy-paste examples in my language

   weight 2 · round to Stripe
   Stripe✓fullcommunity8/10

   Stripe's docs offer copy-paste checkout/payment-link/Elements integrations, sandbox testing with test cards, idempotency, webhooks, and API keys, and community testimony confirms fast integration ('up and running and accepting recurring payments in less than an hour', clear documentation) despite some unrelated support gripes. missing for 10: explicit multi-language SDK code-sample evidence in the pack (only generic doc references, not shown per-language snippets) and a broader/more recent independent corroboration beyond one old HN thread.

   • [claimed-docs] “Build a fully customized payment page using elements”
   • [claimed-docs] “Customers enter their payment details in a fully-featured payment page, either embedded on your site or via a redirect to a Stripe-hosted pa…”
   • [claimed-docs] “Test your integration in a sandbox ... by simulating transactions with test values—these transactions don't move funds.”
   • [claimed-docs] “Test cards act as "fake" credit cards, and allow you to simulate the following scenarios: - Successful payments... - Card errors due to decl…”
   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”
   • [claimed-docs] “You can create an HTTPS webhook endpoint to receive events.”
   • [community] “Stripe is a game changer. I've been using it for a few months and honestly its the best API I've ever used. The documentation is clear and c…”

   proof ↗⚑ flag

   Checkout.com~partialprobed5/10

   Checkout.com's docs support several low-code paths (Hosted Payment Page explicitly marketed as 'less than a day', Flow, Payment Links, plugins) and there's a Unified Payments API plus a public Node.js SDK on GitHub, suggesting a developer could get a basic flow running quickly. However, the pack shows only one language SDK (Node.js) with no evidence of multi-language copy-paste snippets in the docs themselves, and openapi/reference-doc probes returned 404s, weakening confidence that a full in-docs multi-language integration experience exists. missing for 10: multi-language code samples embedded in docs, other official SDKs (Python/Java/PHP/Ruby), independent hands-on confirmation of 'under a day' integration.

   • [claimed-docs] “Get a fully branded checkout page in less than a day. No heavy lifting or confusing compliance. Just one integration, and you're ready to ta…”
   • [claimed-docs] “Flow is a clever bit of code that helps you boost conversions, stay compliant, and enter new markets fast. Accept payments using customizabl…”
   • [claimed-docs] “Take complete control of the entire payment lifecycle from a single API. Upgrade, expand, and adapt your payment experiences with a flexible…”
   • [github] “The official Node.js SDK for Checkout.com payment gateway.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://www.checkout.com/openapi.json, https://www.checkout.com/swagger.json, https://www.checkout.c…”
   • [claimed-docs] “Customize your Hosted Payments Page”

   proof ↗⚑ flag

Sandbox fidelity

1. developerBuild and test the entire payment flow in a sandbox or test mode that mirrors production, with test cards for every scenario including declines and disputes

   weight 3 · round to Stripe
   Stripe✓fullclaimed9/10

   Stripe provides a documented sandbox environment mirroring production (isolated per-team sandboxes, no live-data impact) plus an extensive test-card matrix covering successes, declines, fraud, and invalid data, and a full disputes API for simulating/handling chargebacks. CLI and Workbench also support building/testing/debugging integrations end-to-end. Missing for 10: no independent/hands-on developer testimonial specifically validating sandbox fidelity or test-card coverage beyond first-party docs.

   • [claimed-docs] “Test your integration in a sandbox ... by simulating transactions with test values—these transactions don't move funds.”
   • [claimed-docs] “A sandbox is an isolated test environment. You can use your sandbox to test Stripe functionality in your account, and experiment with new fe…”
   • [claimed-docs] “You can use your sandbox to test Stripe functionality in your account, and experiment with new features without affecting your live integrat…”
   • [claimed-docs] “You can invite another user, such as an implementation partner or design agency, to access all sandboxes, or a specific sandbox, without pro…”
   • [claimed-docs] “Your team can test in separate sandboxes to make sure that data and actions are completely isolated from other sandboxes.”
   • [claimed-docs] “Test cards act as "fake" credit cards, and allow you to simulate the following scenarios: - Successful payments... - Card errors due to decl…”
   • [claimed-docs] “With the API, you can upload evidence, respond to disputes, and receive dispute events using webhooks.”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”
   • [claimed-docs] “Workbench provides powerful tools that help you build, test, and debug your Stripe integration.”
   • [claimed-docs] “Build, test, and manage your Stripe integration from the command line.”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Checkout.com documents a dedicated test-account environment with a public 'Get test account' signup, plus per-scenario testing docs covering payments testing, test cards, disputes testing, fraud-detection testing, and bank payouts testing — directly matching the story's requirement for a sandbox mirroring production with test cards for declines and disputes. Missing for 10: independent/hands-on developer corroboration that the sandbox truly mirrors production behavior end-to-end, and more detail on how comprehensive the decline-scenario coverage is beyond the docs listing.

   • [claimed-docs] “A dedicated test-account environment with per-scenario testing docs: payments testing, test cards, disputes testing, fraud-detection testing…”
   • [claimed-docs] “Test cards”
   • [claimed-docs] “Disputes testing”

   proof ↗⚑ flag

Ai disputes

1. ai-native userHave an agent handle a dispute end-to-end, retrieving it, assembling evidence, and submitting the response via the API

   weight 3 · round to Stripe
   Stripe✓fullclaimed7/10

   Stripe docs explicitly state the Disputes API supports retrieving, uploading evidence, responding to disputes, and receiving dispute webhook events, and the hosted MCP server's documented tool surface includes 'disputes (update)' alongside generic read/write API access, restricted-key scoping for agents, and webhook-based event delivery — covering the retrieve→assemble→submit flow programmatically. Missing for 10: a concrete end-to-end worked example (e.g., a sample agent script that assembles and submits evidence), and clarity on whether dispute submission falls under the human-confirmation gate applied to other write actions.

   • [claimed-docs] “With the API, you can upload evidence, respond to disputes, and receive dispute events using webhooks.”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “Receiving webhook events helps you respond to asynchronous events, such as when a customer's bank confirms a payment, a customer disputes a …”
   • [claimed-docs] “Restricted API keys are Stripe's documented least-privilege credential for agents: "When you create a RAK in the Stripe Dashboard, you selec…”

   proof ↗⚑ flag

   Checkout.com—none0/10

   Checkout.com's docs cover general dispute management topics (manage-disputes, disputes testing, preventing fraudulent disputes) but there's no evidence of an API endpoint for retrieving disputes or submitting evidence, and the documented MCP server's capabilities explicitly list only payment status queries, refunds, voids, and payment links — dispute handling is not among them. No evidence shows an agent can retrieve a dispute, assemble evidence, and submit a response end-to-end via API or MCP.

   • [claimed-docs] “Preventing fraudulent disputes”
   • [claimed-docs] “Disputes testing”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”

   proof ↗⚑ flag

Disputes

1. finance-ops userRespond to a chargeback with structured evidence and track the dispute lifecycle to resolution

   weight 3 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe's Disputes API lets finance-ops programmatically manage disputes, upload structured evidence, respond to disputes, and receive dispute lifecycle events via webhooks, plus dispute events surface in the Dashboard/Radar for tracking; MCP tooling also exposes a 'disputes: update' method for agent-assisted workflows. missing for 10: no independent/hands-on account of the full evidence-submission UX or explicit dispute-resolution outcome tracking beyond webhook events.

   • [claimed-docs] “With the API, you can upload evidence, respond to disputes, and receive dispute events using webhooks.”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”
   • [claimed-docs] “Receiving webhook events helps you respond to asynchronous events, such as when a customer's bank confirms a payment, a customer disputes a …”
   • [claimed-docs] “Use Radar to evaluate transactions, accounts, and customers in real time, using AI algorithms to assess the risk of fraud.”
   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed4/10

   Checkout.com documents a disputes management area ("Preventing fraudulent disputes"), dedicated disputes testing scenarios, and webhook events that report "progress on disputes," showing the platform supports dispute lifecycle tracking. However, there is no evidence of the actual structured-evidence submission workflow (e.g., uploading proof documents, dispute reason codes, response deadlines) or a documented end-to-end resolution tracking UI/API. Missing for 10: documentation of evidence submission fields/API, dispute status states and resolution outcomes, and any dashboard/API walkthrough of responding to a chargeback.

   • [claimed-docs] “Preventing fraudulent disputes”
   • [claimed-docs] “Disputes testing”
   • [claimed-docs] “Create risk rules and lists”
   • [claimed-docs] “Retrieve fraud reports”
   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”

   proof ↗⚑ flag

Fraud prevention

1. finance-ops userScreen transactions with machine-learning fraud scores and write custom block, review, and allow rules

   weight 3 · round to Checkout.com
   Stripe~partialclaimed5/10

   Stripe Radar is documented to use AI/ML algorithms in real time to assess fraud risk across transactions, accounts, and customers, which covers the ML fraud-scoring part of the story; however, the evidence pack does not show custom rule-writing capability (block/review/allow rules) beyond the single Radar mention, no details on rule builder UI, thresholds, or list management. missing for 10: documentation of custom rule creation/editing (block, review, allow lists), rule logic examples, and independent/hands-on confirmation of Radar's rule-writing workflow.

   • [claimed-docs] “Use Radar to evaluate transactions, accounts, and customers in real time, using AI algorithms to assess the risk of fraud.”

   proof ↗⚑ flag

   Checkout.com~partialclaimed6/10

   Checkout.com's docs explicitly cover 'Create risk rules and lists' and 'Retrieve fraud reports' under Prevent Fraud, giving finance-ops the ability to write custom block/review/allow-style rules, and it separately touts an 'AI-driven solution' for optimizing acceptance. However, the evidence never names a machine-learning fraud score, a scoring model, or how scores feed into rule conditions, nor a review-queue UI. Missing for 10: explicit ML fraud-score documentation, rule-to-score integration details, and hands-on/independent confirmation of the scoring engine's accuracy or usability.

   • [claimed-docs] “Create risk rules and lists”
   • [claimed-docs] “Retrieve fraud reports”
   • [claimed-docs] “Preventing fraudulent disputes”
   • [claimed-docs] “Automatically optimize payments to increase your acceptance rate with our powerful AI-driven solution.”

   proof ↗⚑ flag

Refunds

1. finance-ops userIssue full or partial refunds from both the dashboard and the API

   weight 3 · round to Stripe
   Stripe✓fullclaimed9/10

   Docs explicitly state refunds can be issued via both the Refunds API and the Dashboard, including full or partial amounts, and refunded payments are viewable in the Dashboard. No-code refunds are also supported on Payment Links. missing for 10: no independent/hands-on confirmation of partial-refund UX in the dashboard.

   • [claimed-docs] “You can cancel a payment before it's completed at no cost. Or you can refund all or part of a payment after it succeeds, which might incur a…”
   • [claimed-docs] “You can issue refunds by using the Refunds API or the Dashboard”
   • [claimed-docs] “You can issue refunds by using the Refunds API or the Dashboard.”
   • [claimed-docs] “You can view your refunded payments in the Dashboard”
   • [claimed-docs] “No-code refunds Yes”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Docs explicitly cover refunding a payment via API with or without a reference (checkout-com-docs-21, checkout-com-docs-55), implying support for full/partial refunds through the payments/{id}/refunds endpoint (also listed as idempotent, cko-supp-idempotency). Dashboard-based refund management is implied by broader webhook/dashboard management patterns but not explicitly documented with a dedicated dashboard refund walkthrough. Missing for 10: explicit dashboard UI screenshot/walkthrough for issuing refunds, and independent/hands-on confirmation of partial refund amounts being configurable.

   • [claimed-docs] “Refund a payment without a reference”
   • [claimed-docs] “Refund a payment with a reference”
   • [claimed-docs] “Idempotent retries are first-class: "The following endpoints from our API support idempotent requests, so you can safely retry them without …”

   proof ↗⚑ flag

Multi currency

1. founderCharge customers in their local currencies and settle to my bank in mine with transparent FX handling

   weight 2 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe documents charging in 135+ currencies with payout settlement in a merchant's preferred currency (stripe-docs-19), localized checkout/payment links that let customers pay in their own currency (stripe-docs-36), automated payout tracking to bank accounts (stripe-docs-34, stripe-docs-52, stripe-docs-68), and automated tax compliance across jurisdictions (stripe-docs-20, stripe-docs-48). This directly covers charging locally and settling in the founder's currency. Missing for 10: explicit documentation of FX conversion rate transparency/fee disclosure mechanics and independent/hands-on confirmation of FX rate fairness or settlement accuracy.

   • [claimed-docs] “You can charge customers in over 135 currencies and receive funds in your preferred currency.”
   • [claimed-docs] “Let your customers use a link to pay in their local language and currency with their preferred payment method.”
   • [claimed-docs] “Stripe sends funds from your available balance to your bank account as payouts”
   • [claimed-docs] “Track a payout: View your payouts and their expected deposit dates in the Dashboard.”
   • [claimed-docs] “Set up your bank account to receive payouts.”
   • [claimed-docs] “Automate sales tax, VAT, and GST compliance on all your transactions.”
   • [claimed-docs] “Collect tax on Payment Links: Learn how to calculate and collect tax on a payment page without writing any code.”

   proof ↗⚑ flag

   Checkout.com~partialclaimed6/10

   Checkout.com clearly supports charging in local currencies (150+ currencies, local acquiring, FX rates API for acquiring/payouts) and provides settlement reporting and a documented FX rate fee for reconciliation, which supports the founder's need to settle in their own currency. However, evidence does not show a clear, transparent published FX markup/rate table or a self-serve explanation of exactly how conversion and settlement currency selection works — pricing and FX fees are sales-led/unpublished, undercutting the 'transparent' claim. Missing for 10: a documented transparent FX rate/markup disclosure mechanism, self-serve settlement currency selection docs, and independent verification of actual FX transparency in practice.

   • [claimed-docs] “Use the FX rates API to retrieve the indicative foreign exchange (FX) rates that Checkout.com uses for acquiring and card payouts.”
   • [claimed-docs] “Process worldwide in 150+ currencies. Plus domestic coverage in 45+ countries and counting.”
   • [claimed-docs] “Accept payments like a local business globally. Get paid in local currencies, settle funds fast, and gain valuable insights.”
   • [claimed-docs] “"You can reconcile your balances, fee categories, financial actions, invoices, and settlements with Checkout.com using the following reports…”
   • [claimed-docs] “Exchange (FX) rate fee”
   • [claimed-docs] “Pricing is sales-led and unpublished: a tailored "Fully Flat-Rate" plan ("We price based on your business profile and risk category, nothing…”
   • [claimed-docs] “Settlement Breakdown Report”

   proof ↗⚑ flag

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

   weight 2 · round to Stripe
   Stripe✓fullclaimed8/10

   Stripe's core resources (payments, refunds, subscriptions, invoices, disputes, payouts, tax, Connect) are all documented as available via the REST API with idempotency, versioning, and restricted keys, and the MCP server explicitly exposes generic stripe_api_read/stripe_api_write tools plus task-specific tools mirroring Dashboard actions across most objects. Workbench Shell/CLI further lets agents run the same commands as the CLI/API interactively. missing for 10: independent confirmation that every Dashboard-only feature (e.g., third-party Stripe Apps like Kyber no-code automations) has a full API equivalent, and note that sensitive stripe_api_write actions require human confirmation rather than being fully autonomous.

   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “With the API, you can upload evidence, respond to disputes, and receive dispute events using webhooks.”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”
   • [claimed-docs] “The API supports idempotency for safely retrying requests without accidentally performing the same operation twice.”
   • [claimed-docs] “Each monthly release includes only backward-compatible changes... You can safely upgrade to a new monthly release without breaking any exist…”
   • [claimed-docs] “A restricted API key (RAK) is an API key ... a RAK can do only what you give it permission to do.”
   • [claimed-docs] “Workbench (Dashboard > Developers > Workbench) ships an interactive Shell and API Explorer: "Shell is a command line interface within Workbe…”
   • [claimed-docs] “Build, test, and manage your Stripe integration from the command line.”
   • [claimed-docs] “Docs, "Confirm actions by agents acting on your behalf": "To prevent agents from making mistakes, Stripe requires human confirmation before …”

   proof ↗⚑ flag

   Checkout.com~partialprobed6/10

   Checkout.com's core payment lifecycle (payments, refunds, voids, payment links, webhooks, reconciliation, platforms/sub-entities, FX) is fully API-driven per docs, and the API is treated as the primary integration surface rather than the UI. However, some Dashboard-only or UI-centric functions (e.g., some fraud rule configuration, risk list management, business account setup, certain reporting views) are only shown via Dashboard screenshots/flows without confirmed API parity, and no OpenAPI spec was discoverable (404s on probes) to verify full API surface coverage. missing for 10: confirmed API parity for all Dashboard-configurable settings (risk rules, fraud lists, business account management), a public/complete OpenAPI spec, and independent evidence confirming no UI-only features exist.

   • [claimed-docs] “Unified Payments API: Connect to all payment methods through a single API integration.”
   • [claimed-docs] “Idempotent retries are first-class: "The following endpoints from our API support idempotent requests, so you can safely retry them without …”
   • [claimed-docs] “Two documented server-side authentication forms — "Access keys (OAuth 2.0)" (client-credentials flow) and "Secret keys" — plus public keys f…”
   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “"You can reconcile your balances, fee categories, financial actions, invoices, and settlements with Checkout.com using the following reports…”
   • [claimed-docs] “"The Platforms solution enables marketplaces, payment facilitators (Payfacs), and other platform-based businesses to onboard sub-entities an…”
   • [claimed-docs] “Create risk rules and lists”
   • [claimed-docs] “Retrieve fraud reports”
   • [probe] “PROBE openapi: all candidate paths 404 (https://www.checkout.com/openapi.json, https://www.checkout.com/swagger.json, https://www.checkout.c…”
   • [claimed-docs] “Refund a payment without a reference”
   • [claimed-docs] “Refund a payment with a reference”

   proof ↗⚑ flag

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

   weight 3 · round drawn
   Stripe~partialclaimed3/10

   Evidence shows only narrow CSV export of specific financial reports (balance, payouts, activity, tax, Revenue Recognition) via stripe_report/Workbench, not a comprehensive account data export mechanism for a user to leave the platform. Missing for 10: a documented full account/data export tool (customers, charges, subscriptions, full transaction history) in open/portable formats, and any explicit data-portability or account-closure export workflow.

   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”
   • [claimed-docs] “Use these files to generate SDKs or client libraries that support Stripe's API.”

   proof ↗⚑ flag

   Checkout.com~partialclaimed3/10

   Checkout.com documents various financial reports (Settlement Breakdown, Balance, Payouts, Reconciliation) accessible via Dashboard/API, which implies some data export capability, but there is no explicit documentation of a comprehensive 'export all your data in open formats and leave' capability, no data portability policy, and no mention of open-format guarantees or full account data export upon departure. missing for 10: explicit full-data export tooling, open-format (CSV/JSON) export guarantees, documented data portability/account closure export process, and independent confirmation of completeness.

   • [claimed-docs] “Reconciliation is a crucial part of your business's bookkeeping process. You can reconcile your balances, fee categories, financial actions,…”
   • [claimed-docs] “Settlement Breakdown Report”
   • [claimed-docs] “"You can reconcile your balances, fee categories, financial actions, invoices, and settlements with Checkout.com using the following reports…”

   proof ↗⚑ flag

Ai reconciliation

1. ai-native userHave an agent pull payout and balance-transaction data via the API and reconcile each bank deposit to its underlying charges

   weight 3 · round to Checkout.com
   Stripe~partialclaimed7/10

   Stripe's API and MCP tooling explicitly expose payout and balance-transaction data for programmatic/agent access — the documented stripe_api_read method list includes 'payouts, and balance transactions', and stripe_analytics/stripe_report can run balance and payout financial reports, letting an agent pull the data needed to match deposits to charges (stripe-supp-mcp-tools, stripe-supp-analytics-ai, stripe-docs-34/52/68). However, there is no first-party tutorial or worked example specifically demonstrating automated reconciliation logic (matching a bank deposit to its underlying charges), and no independent/hands-on evidence confirming this reconciliation workflow works end-to-end. Missing for 10: a documented reconciliation workflow/tutorial tying payout amounts to constituent balance-transaction charges, and independent verification of an agent successfully performing this reconciliation.

   • [claimed-docs] “The hosted MCP server (mcp.stripe.com) exposes generic API tools plus task tools: `stripe_api_search` ("Search for Stripe API methods by key…”
   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”
   • [claimed-docs] “Stripe sends funds from your available balance to your bank account as payouts”
   • [claimed-docs] “Track a payout: View your payouts and their expected deposit dates in the Dashboard.”
   • [claimed-docs] “Set up your bank account to receive payouts.”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Checkout.com documents a report suite explicitly designed for reconciling payouts to underlying financial actions (Balance Breakdown, Balance, Financial Actions by Date Range/Payout ID, Payouts, Settlement Breakdown, Settlement Statement reports), and these are exposed via API alongside a documented MCP server and OAuth2/API-key auth suitable for agent-driven automation. Missing for 10: no explicit end-to-end example showing an agent script pulling payout+balance-transaction endpoints and matching them to charges, and no independent/hands-on confirmation of this reconciliation workflow being done via an agent.

   • [claimed-docs] “"You can reconcile your balances, fee categories, financial actions, invoices, and settlements with Checkout.com using the following reports…”
   • [claimed-docs] “Reconciliation is a crucial part of your business's bookkeeping process. You can reconcile your balances, fee categories, financial actions,…”
   • [claimed-docs] “Settlement Breakdown Report”
   • [claimed-docs] “Two documented server-side authentication forms — "Access keys (OAuth 2.0)" (client-credentials flow) and "Secret keys" — plus public keys f…”
   • [claimed-docs] “The Checkout.com MCP Server (Beta) "enables you to search through our knowledge base and manage payment operations directly from your AI-ass…”
   • [claimed-docs] “The MCP server is hosted with documented per-environment endpoints — "Sandbox: https://checkout.mcp.sbox.cko.tech Production: https://mcp.ch…”

   proof ↗⚑ flag

Payouts

1. finance-ops userControl the payout schedule and see exactly which charges, fees, and refunds make up each payout

   weight 2 · round drawn
   Stripe~partialclaimed6/10

   Stripe documents payout basics (funds sent from balance to bank account, dashboard tracking of payouts and expected deposit dates, setting up bank accounts) and financial reporting via stripe_report/analytics (balance, payouts, activity reports downloadable as CSV), which supports reconciliation. However, evidence does not show explicit payout schedule controls (e.g., setting manual vs automatic schedule, daily/weekly/monthly cadence) nor a documented per-payout breakdown view showing exact charges/fees/refunds composing each payout. missing for 10: explicit payout schedule configuration docs, a documented itemized payout reconciliation report/API showing constituent charges/fees/refunds per payout, independent/hands-on confirmation of this reconciliation workflow.

   • [claimed-docs] “Stripe sends funds from your available balance to your bank account as payouts”
   • [claimed-docs] “Track a payout: View your payouts and their expected deposit dates in the Dashboard.”
   • [claimed-docs] “Set up your bank account to receive payouts.”
   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”

   proof ↗⚑ flag

   Checkout.com~partialclaimed6/10

   Checkout.com documents a strong reconciliation report suite mapping payouts to their underlying financial actions — the 'Financial Actions by Payout ID Report', 'Payouts Report', 'Settlement Breakdown Report', and 'Balance' reports explicitly let finance-ops trace charges, fees, and refunds composing a payout, plus FX fee visibility. However, there is no evidence of the 'control the payout schedule' half of the story (e.g., configuring payout frequency/cadence or triggering ad hoc payouts) anywhere in the pack. Missing for 10: documentation of payout schedule/frequency configuration, self-serve payout timing controls.

   • [claimed-docs] “"You can reconcile your balances, fee categories, financial actions, invoices, and settlements with Checkout.com using the following reports…”
   • [claimed-docs] “Settlement Breakdown Report”
   • [claimed-docs] “Exchange (FX) rate fee”
   • [claimed-docs] “Reconciliation is a crucial part of your business's bookkeeping process. You can reconcile your balances, fee categories, financial actions,…”

   proof ↗⚑ flag

Reconciliation

1. finance-ops userReconcile settlements to bank deposits with itemized reports covering fees, refunds, and chargebacks

   weight 3 · round to Checkout.com
   Stripe~partialclaimed5/10

   Stripe documents payouts to bank accounts with expected deposit dates (stripe-docs-34, stripe-docs-52), refunds and dispute management via API/Dashboard (stripe-docs-17, stripe-docs-33, stripe-docs-41), and AI-assisted analytics/report runs including a documented 'balance summary' report and other financial report types (stripe-supp-analytics-ai). However, there is no evidence of a purpose-built, itemized settlement-to-deposit reconciliation report that breaks out fees, refunds, and chargebacks against a specific bank deposit in one view — the closest evidence (report types: balance, payouts, activity, tax) implies this exists but doesn't confirm itemization matching finance-ops reconciliation workflows, and third-party accounting sync (QuickBooks, Xero) is only listed as partner integrations rather than native itemized reconciliation. missing for 10: a documented native 'payout reconciliation report' itemizing fees/refunds/chargebacks per deposit, confirmation of automatic matching to bank statement lines, and independent/hands-on evidence from finance-ops users that this reconciliation workflow works well in practice.

   • [claimed-docs] “Stripe sends funds from your available balance to your bank account as payouts”
   • [claimed-docs] “Track a payout: View your payouts and their expected deposit dates in the Dashboard.”
   • [claimed-docs] “You can cancel a payment before it's completed at no cost. Or you can refund all or part of a payment after it succeeds, which might incur a…”
   • [claimed-docs] “You can issue refunds by using the Refunds API or the Dashboard”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”
   • [claimed-docs] “Agents can analyze aggregated account data through MCP: `stripe_analytics` "handles three kinds of requests: Metric queries for Stripe-defin…”
   • [claimed-docs] “We are specialists in connecting Stripe to QuickBooks Online”
   • [claimed-docs] “View relevant Xero customer and transaction data within the Stripe Dashboard”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Checkout.com documents a dedicated reconciliation report suite (Balance Breakdown, Balance Report, Financial Actions by Date Range/Payout ID, Payouts Report, Settlement Breakdown Report, Settlement Statement) explicitly mapping settlements to underlying financial actions including fee categories, and separate refund/dispute docs cover those transaction types. Missing for 10: independent/third-party corroboration (e.g. finance-ops user reviews) that reconciliation reports itemize chargebacks specifically and match to bank deposits in practice, and no sample report output shown.

   • [claimed-docs] “"You can reconcile your balances, fee categories, financial actions, invoices, and settlements with Checkout.com using the following reports…”
   • [claimed-docs] “Reconciliation is a crucial part of your business's bookkeeping process. You can reconcile your balances, fee categories, financial actions,…”
   • [claimed-docs] “Settlement Breakdown Report”
   • [claimed-docs] “"Webhooks are automated messages that Checkout.com sends your server when events related to your account occur. For example, payment lifecyc…”
   • [claimed-docs] “Refund a payment without a reference”
   • [claimed-docs] “Refund a payment with a reference”
   • [claimed-docs] “Preventing fraudulent disputes”
   • [claimed-docs] “Exchange (FX) rate fee”

   proof ↗⚑ flag

Marketplace

1. developerBuild a marketplace or platform that onboards sellers, splits payments between parties, and pays out to each seller

   weight 2 · round to Stripe
   Stripe✓fullclaimed9/10

   Stripe Connect is explicitly documented as the product for building marketplaces/platforms that onboard sellers, split payments, and pay out to multiple parties, backed by supporting payouts, payment collection, refunds, disputes, and tax infrastructure. Missing for 10: independent hands-on evidence specifically of a marketplace/Connect integration (community evidence covers Stripe generally, not Connect specifically).

   • [claimed-docs] “Collect payments from customers and automatically pay out a portion to sellers or service providers on your marketplace.”
   • [claimed-docs] “Use Connect to build a platform, marketplace, or other business that manages payments and moves money between multiple parties.”
   • [claimed-docs] “Stripe sends funds from your available balance to your bank account as payouts”
   • [claimed-docs] “Track a payout: View your payouts and their expected deposit dates in the Dashboard.”
   • [claimed-docs] “Set up your bank account to receive payouts.”
   • [claimed-docs] “You can issue refunds by using the Refunds API or the Dashboard”
   • [claimed-docs] “You can programmatically manage disputes using the API. With the API, you can upload evidence, respond to disputes, and receive dispute even…”

   proof ↗⚑ flag

   Checkout.com✓fullclaimed8/10

   Checkout.com has a dedicated 'Platforms' product explicitly for marketplaces/payfacs that documents onboarding sub-entities, splitting funds between platform and sub-entities, and paying out sub-entities in their local currency/preferred method, backed by reconciliation and settlement reporting to track those splits and payouts. missing for 10: independent/hands-on validation of the onboarding flow, detailed KYC/sub-entity API reference specifics, and public case studies of marketplaces running this in production.

   • [claimed-docs] “"The Platforms solution enables marketplaces, payment facilitators (Payfacs), and other platform-based businesses to onboard sub-entities an…”
   • [claimed-docs] “The Platforms solution enables marketplaces, payment facilitators (Payfacs), and other platform-based businesses to onboard sub-entities and…”
   • [claimed-docs] “Split funds between your platform and sub-entities according to your business model.”
   • [claimed-docs] “Make payouts to your sub-entities in their local currency and preferred payment method.”
   • [claimed-docs] “"You can reconcile your balances, fee categories, financial actions, invoices, and settlements with Checkout.com using the following reports…”
   • [claimed-docs] “Settlement Breakdown Report”

   proof ↗⚑ flag

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

   weight 2 · round drawn
   Stripe—none0/10

   No evidence in the pack discusses data residency, regional data storage choices, or GDPR/EU data localization options for Stripe accounts; the evidence covers payments, billing, MCP/agents, sandboxes, etc. but never data residency controls.

   ⚑ flag

   Checkout.com—none0/10

   No evidence in the pack mentions data residency, regional data storage options, or user/customer control over where data is stored; while the product processes payments across many currencies/countries, this does not address data residency choice. This axis is applicable to a payments platform handling sensitive data but no supporting evidence exists.

   ⚑ flag