Adyen MCP server

# Fix list: Adyen MCP server

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/adyen-mcp-server, the October 2026 research run, assessed 9 October 2026. Grade C, 60.3 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 Adyen MCP server: 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. Payments & pricing, 35 out of 100, up to 8.1 more on the total

Why it scored 35: The server is free, so pricing is read on the Adyen account it needs. No x402, MPP or L402 in the docs page, the repository or the pricing page (0). The pricing page lists per-transaction fees without a login, $0.13 plus the payment method's fee, with no setup or monthly fee (20). A developer test account is free through a signup form. The form's fields are drawn by script, so whether it asks for a card was not read, and a minimum invoice applies to live accounts (15 of 20). A person has to sign up, and live accounts go through sales (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.

## 2. Agent ergonomics, 56 out of 100, up to 7.2 more on the total

Why it scored 56: The default loads 38 tools (5), and `--tools` takes a comma-separated subset, which the README advises (8 back, 13 of 25). Terminal list tools take `pageNumber` and `pageSize` and filters. Other tools return the raw API response as one JSON string with no size control (11 of 20). Failures come back as readable text with the API's message, sanitised since 0.5.0, but without `isError` and with no documented codes (8 of 20). All 38 tools carry read-only, destructive and idempotent hints, checked by a test. No POST tool sends an idempotency key, although the API supports one (13 of 20). The test environment is the default and the checkout tools need two to five fields. The server ships for Node only (11 of 15).

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.

## 3. Reliability, 65 out of 100, up to 7 more on the total

Why it scored 65: Read on the local-software lines, because the server runs on the owner's machine over stdio. It installs from the official npm package `@adyen/mcp`, published from GitHub Actions with a provenance attestation, with Node 18 or later stated (20). The repository has a test workflow on every push and four vitest files of 959 lines. GitHub's API refused us before we read the latest run, so a passing result is unconfirmed (18 of 25). One bug report is open, filed on 7 October 2026 and unanswered on the day, about refunds repeated after a timeout. Earlier issues were answered within three days (19 of 25). Releases are tagged by version with notes on GitHub and a breaking-change category in the release template, but there is no changelog file and 0.6.0 reports itself as 0.4.0 (8 of 15). The README is headed alpha and the version is 0.6.0 (0).

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.

## 4. Security & auth, 63 out of 100, up to 6.5 more on the total

Why it scored 63: The credential is an Adyen API key on a web service user with roles and company or merchant scope, replaceable at any time with 24 hours of overlap. The docs page still shows the key on the command line, and 0.6.0 added the environment variable and a warning for live use (26 of 30). `--tools` and roles narrow what the server can do, the test environment is the default, and writes are annotated. No read-only switch and no confirmation step for refunds or terminal changes was found (11 of 20). Tools return webhook settings, user records and payment data written by other people, and no guidance on untrusted content was found (3 of 15). The Customer Area keeps API logs of each request and response for 30 days (12 of 15). Adyen has a responsible disclosure policy with a 60-day fix target and a hall of fame, its terms commit to PCI DSS certification, and the repository runs CodeQL. security.txt expired on 31 December 2025, the repository has no security policy file, and no bug bounty was found (11 of 20).

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.

## 5. Maintenance & community, 42 out of 100, up to 5.1 more on the total

Why it scored 42: 0.6.0 was published on 24 August 2026, 46 days before this check (20 of 30). It is the only release in the last 90 days, with 0.5.0 on 22 April (0). Maintainers answered two outside issues within one and three days in September and merged dependency updates on 11 September. Two outside pull requests from July were still open, and the 7 October bug report had no reply after two days (15 of 25). The repository has no `server.json` or `mcpName`, and the one page of the official registry we could read listed only a third party's Adyen server (0). Renovate, CodeQL, lint and test workflows run, and npm publishing uses provenance. `@adyen/api-library` is pinned at version 27 while an update to 32 waits (7 of 10).

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.

## 6. Schema & documentation, 73 out of 100, up to 4.4 more on the total

Why it scored 73: Every tool is registered with a zod shape that the MCP SDK turns into JSON Schema (25). docs.adyen.com has an llms.txt and a Markdown twin of every page, the MCP page included (10). Descriptions give arguments, returns, notes and an example, and several say when to pick another tool. Some are wrong. `refund_payment` names `paymentPspReference` where the schema has `pspReference` and omits the required `reference`, and `cancel_payment` points to reversal tools that do not exist (12 of 20). Terminal tools have enums and described fields. The checkout tools take currency, country and status as free strings (9 of 15). Each description has an example. No error reference for the tools was found, though Adyen documents HTTP status and error codes for the APIs (8 of 15). Releases are versioned with notes on GitHub, while the docs page was last modified on 11 June 2025 and names two of the four APIs covered (9 of 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.

## 7. Transparency & trust, 83 out of 100, up to 1.5 more on the total

Made of editorial 72, provenance 93.

Why it scored 83: The server is MIT on GitHub (30). It runs locally and sends requests only to Adyen's APIs. Adyen's privacy statement of 4 August 2025 and a sub-processor list of 1 July 2026 are public. We did not read retention periods, and nothing addresses MCP use (20 of 30). Adyen documents API versioning and what counts as a breaking change. The server has no deprecation policy and is labelled alpha (8 of 20). No telemetry code was found in the source. The server sets the application name `adyen-mcp-server 0.4.0` on its API requests, which the README does not mention (14 of 20).

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.

Provenance checks not met in full (half of this category, computed from checked facts):

- Terms of service: read, states 7 of the 7 things a reader expects, and has 1 clause that costs points (8 of 10)
- security.txt: published but past its Expires date (5 of 10)

## 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: the result of the latest CI run on the default branch, because GitHub's API rate limit refused the request
- unchecked: whether an Adyen-namespaced entry exists in the official MCP registry. One search page was read and three later requests timed out or errored. The repository has no `server.json`
- unchecked: the status page's component list and incident history, which are drawn by script
- unchecked: the fields of the test account signup form, drawn by script, so whether it asks for a card
- unchecked: retention periods in the privacy statement, Adyen's certifications beyond the PCI DSS line in the terms, and any bug bounty
- unchecked: published API rate limit numbers. None were found in the docs index, which was searched by title only
- The lead said the tools cover the Checkout and Management APIs. The source also has Legal Entity Management tools and `get_account_holder` on the Balance Platform configuration API
- The pricing page shows payment method names as images, so the 0.60% markup could not be tied to named methods in the text we read
- Whether Adyen will change `refund_payment` after issue 149. A September 2026 issue on retry safety was closed as not an issue

## Weaknesses

- The README is headed alpha, the package is at 0.6.0, and the server reports itself as version 0.4.0
- `refund_payment`, `cancel_payment` and `create_payment_links` send no `idempotency-key` header, although Adyen's API accepts one on POST requests
- Tool failures return as text such as `Failed to refund payment. Error: ...` without `isError`, and no tool reads a payment's refunds
- The `refund_payment` description names `paymentPspReference` while the schema field is `pspReference`, and omits the required `reference`
- One release in the 90 days to 9 October 2026, and the docs page, last modified 11 June 2025, still passes the key on the command line

## 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.

- Set `ADYEN_API_KEY` in the environment and start with `npx -y @adyen/mcp --env=TEST`. The `--adyenApiKey` flag in the docs page puts the key in the process list
- Pass `--tools=` with only the tool names the task needs. The default loads all 38, and the terminal settings schema alone is about 29 KB of source
- Call `refund_payment` with `pspReference`, `currency`, `value`, `merchantAccount` and `reference`. The description's `paymentPspReference` is not the field name
- Treat a `Failed to ...` reply from a write tool as outcome unknown, not as failure. Do not repeat a refund without checking the Customer Area or the REFUND webhook
- Give amounts in minor units (1099 for 10.99 EUR). Live use needs `--env=LIVE --livePrefix=` with the prefix from the live Customer Area

## 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.