Stripe API + MCP by Stripe

# Fix list: Stripe API + MCP

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/stripe-mcp, the October 2026 research run, assessed 1 October 2026. Grade A, 82.4 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Stripe API + MCP: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Reliability, 63 out of 100, up to 7.4 more on the total

Why it scored 63: Status page at status.stripe.com describes itself as real-time and historical data (20), but it renders only in JavaScript and our reader got "Loading...", so we couldn't read the last 90 days (5). Rate limits published, 100 requests a second in live mode, 25 in a sandbox, 25 per endpoint, plus per-resource limits (15). Every 429 carries a `Stripe-Rate-Limited-Reason` header, the docs prescribe exponential backoff with jitter, lock timeouts are retried by the SDKs, and the API takes idempotency keys with a dedicated `idempotency_error` (15). The pricing page cites 99.999% average historical uptime, which is a record rather than a commitment, and we found no SLA (0). The API, MCP server and machine payments are live, while `stripe_analytics` and the Treasury balance tool are marked preview (8 of 10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 2. Payments & pricing, 65 out of 100, up to 4.4 more on the total

Why it scored 65: Payment platforms and wallets take the highest step that applies on the 40-point protocol line. 40 for x402, MPP or L402 on all their own endpoints, 30 on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, 0 for only a protocol of their own. Stripe merchants can accept MPP (cards through shared payment tokens, USDC.e on Tempo, USDC on Solana) and x402 (USDC on Base) on their own endpoints, settled to the Stripe balance, while Stripe's own API and MCP server aren't paid over either, so the merchant step (25 of 40). Per-unit prices published without a login, 2.9% plus 30 cents for US cards, 1.5% for stablecoins, $0.15 per shared payment token issued, 0.7% of Billing volume (20). No setup or monthly fees and free sandboxes, so nothing needs a card to start (20). A person creates the Stripe account and connects the MCP client (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 3. Schema & documentation, 90 out of 100, up to 1.6 more on the total

Why it scored 90: Public OpenAPI 3 spec in stripe/openapi (MIT), 431 paths, regenerated for API version 2026-09-30.endive on 30 September (25). llms.txt and every page as Markdown, also readable through `stripe docs` in the CLI (10). The MCP page describes each of the ten tools and warns about prompt injection when mixing servers, though the generic read and write tools leave method choice to the agent (15). The API is fully typed with enums, but `stripe_api_write` takes any POST, PATCH, PUT or DELETE method, so the MCP contract is looser than the API's (10). Examples for every endpoint and documented error types (15). Dated API versions pinned per request with `Stripe-Version`, and a public Developer Changelog (15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 4. Maintenance & community, 82 out of 100, up to 1.6 more on the total

Why it scored 82: API version 2026-09-30.endive released on 30 September 2026, and the OpenAPI repo was updated on 1 October (30). Dated API versions and spec updates every few days through the last 90 days (20). Public Developer Changelog, support, and 62 commits to stripe/ai since 1 July; we didn't read the issue queue (14 of 15). com.stripe/mcp is in the official registry, but the entry is version 0.2.4 from 28 October 2025 and points at the old stripe/agent-toolkit repo name (12 of 15). The npm and PyPI packages in stripe/ai haven't had a version bump since May 2026, though CI runs on every pull request with actions pinned to commit SHAs (6).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## 5. Transparency & trust, 87 out of 100, up to 1.1 more on the total

Made of editorial 74, provenance 100.

Why it scored 87: stripe/ai and stripe/openapi are MIT and the service is closed under published terms (18). The security page links the privacy policy, `Privacy Center` and DPA, states a data-retention policy without periods, and cites EU-US, UK and Swiss Data Privacy Framework participation and CBPR and PRP certifications (24). Dated API versions with an upgrade guide, and the MCP key change was announced ahead of its 31 October 2026 date (18). A DPA and data-transfer frameworks are published; we didn't open the subprocessor list (14).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

## 6. Agent ergonomics, 94 out of 100, up to 1 more on the total

Why it scored 94: Ten MCP tools, with `stripe_api_search` and `stripe_api_details` fetching method details on demand instead of loading the whole API into context (25). List endpoints take `limit`, cursor pagination, filters, search and `expand` (20). Errors carry a type, code and message, and rate-limit responses say which limit was hit (20). Idempotency keys on the API and human confirmation for refunds and outbound payments through the MCP server; we couldn't check tool annotations on the hosted server (17). Official SDKs in seven or more languages and the agent toolkit in TypeScript and Python, but the generic tools mean a search, details and write sequence for most actions (12).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 7. Security & auth, 97 out of 100, up to 0.5 more on the total

Why it scored 97: OAuth for interactive MCP clients with per-account and per-environment permissions and revocable sessions, restricted and Agent-tagged API keys with chosen permissions, and API key access policies by location. From 31 October 2026 the MCP server rejects full-access secret keys (30). Least privilege through restricted keys and Dashboard roles, and Stripe requires a person to approve sensitive `stripe_api_write` actions such as refunds and outbound payments, with approvals expiring after 24 hours (20). The MCP page tells users to turn on human confirmation of tools and to take care combining Stripe with other servers because of prompt injection; customer-entered fields still come back through `stripe_api_read` (12). MCP tool-call logs in Workbench and an exportable security history (15). HackerOne bug bounty, PCI Service Provider Level 1, annual SOC 1 and SOC 2 Type II reports, a public SOC 3, and a valid security.txt per the 30 September check (20). Machine payments land in the Stripe balance, so Stripe holds the funds until payout.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: incident history for the last 90 days, because status.stripe.com renders only in JavaScript
- unchecked: readOnlyHint and destructiveHint annotations on the hosted MCP tools
- unchecked: the subprocessor list linked from the DPA
- Whether Stripe will update the registry entry (0.2.4, October 2025) to match the hosted server

## Weaknesses

- Card payments from agents have a 0.50 USD minimum, so per-call micropayments must use stablecoins
- Stablecoin acceptance needs manual approval, excludes New York and is by email request outside the US
- Generic `stripe_api_read` and `stripe_api_write` tools push method choice and parameters onto the agent
- The official MCP registry entry is 0.2.4 from October 2025 and names an old repo
- Status history renders only in JavaScript

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Switch to an Agent-tagged restricted key or OAuth before 31 October 2026; other keys get a 401
- Call `stripe_api_search` and `stripe_api_details` before `stripe_api_write` to get the method and parameters right
- When a write needs approval, give the person the URL and retry only after they approve; approvals expire after 24 hours
- Send an `Idempotency-Key` on every create so a retry can't charge twice
- On 429, read `Stripe-Rate-Limited-Reason` and back off with jitter; a 429 without it is a lock timeout

## What the review panel asked for

- Typed common-action tools
- Status history as JSON
- a registry entry kept in step with the hosted server
- tagged releases for the stripe/ai packages
- Worked agent-payment fee example
- Wider stablecoin access
- Publish the tool descriptions and annotations in the MCP page
- readable status history
- update the registry entry
- A status history agents can read without JavaScript
- Automate stablecoin approval
- published tool annotations

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.