# Fix list: Wikimedia REST API From Anchor Terminal's listing at https://www.anchorterminal.com/tools/wikimedia, the October 2026 research run, assessed 1 October 2026. Grade B, 66.8 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 Wikimedia REST API: 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: Statuspage at wikimediastatus.net with a history feed back to June 2025 (20). Since 3 July, a power failure took services down from 17:26 to 20:10 UTC on 23 September, and Wikipedia and other sites were slow or unavailable from 18:24 to 19:56 UTC on 25 August. A connectivity problem from Russia on 30 and 31 July was outside Wikimedia's control. Two majors sit between one (10) and several (0), so 5 (5). Wikimedia domains are cache-only for our fetcher, so we read the limits from the public mirror of the production deployment config instead. The REST gateway in front of *.wikipedia.org enforces a "May2026" policy per minute by client class, 10 for anonymous clients without a descriptive User-Agent, 200 with one, 200 with OAuth, 2,000 for established users and no practical cap for approved bots. That's numbers in public config, not in docs we could read (10 of 15). Over the limit the gateway returns 429 with a body that links the rate-limit best practices and bot-traffic@wikimedia.org, and MediaWiki maps `ratelimited` and `maxlag` to 429. X-RateLimit headers are off in production and we found no Retry-After (8 of 15). The paid Wikimedia Enterprise plan carries an SLA of 'up to 99%' (10). The v1 REST API on each wiki is generally available, while api.wikimedia.org is being retired (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. Security & auth, 66 out of 100, up to 6 more on the total Why it scored 66: Reads need no credential. Per the 26 September check, OAuth raises limits and is needed for edits. We couldn't load the grant documentation (25 of 30). Anonymous access is read-only, and writes need OAuth (15 of 20). The main payload is article text anyone can edit, and we found no guidance on treating it as untrusted (0 of 15). Every edit is public in page history and user contributions, reads have no log (10 of 15). The production docroot for wikipedia.org serves a security.txt with security@wikimedia.org, a policy link to the reporting process on mediawiki.org, a hall of fame at security.wikimedia.org and an expiry of 2029-03-31. MediaWiki security releases are public. No bug bounty or certification found (16 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. ## 3. Schema & documentation, 68 out of 100, up to 5.2 more on the total Why it scored 68: MediaWiki core ships a discovery endpoint and generated OpenAPI per REST module, and production config sets an availability level per module (published, discoverable, hidden or disabled). The gateway also serves an OpenAPI 3.0.1 file with 46 paths for the older rest_v1 API at `/api/rest_v1/?spec` from the production docroot. We couldn't load either live, since Wikimedia domains are cache-only for our fetcher, and read both in the public config mirrors (25). No llms.txt per the 26 September check, and none in the production docroot (0). Route definitions carry translated descriptions, which we couldn't read rendered (10 of 20). Typed path and query parameters in the route files, not checked against a live spec (10 of 15). The content module declares error schema version 2.0, and we couldn't read examples on mediawiki.org (8 of 15). /v1 paths, per-version release notes and a REST API changelog page. The v2 beta content module is disabled by default in production (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. Agent ergonomics, 68 out of 100, up to 5.2 more on the total Why it scored 68: `/page/{title}/bare` returns metadata without content, and you choose HTML or wikitext per call (18 of 25). Search and page history routes exist, with paging we couldn't confirm in the docs (14 of 20). Errors follow a versioned JSON error schema (14 of 20). Reads are GETs, and edits are separate POST and PUT routes behind OAuth (15 of 20). Few required parameters, and no official SDK for the REST API (7 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. ## 5. Payments & pricing, 60 out of 100, up to 5 more on the total Why it scored 60: No x402, MPP or L402 (0). The REST API is free, and Wikimedia Enterprise publishes a free plan with paid plans by quote (20). Free without a card, including Enterprise's 50,000 on-demand requests a month (20). An agent can call anonymously with only a User-Agent (20). 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. ## 6. Maintenance & community, 70 out of 100, up to 2.6 more on the total Why it scored 70: MediaWiki master took commits up to 28 September 2026 and ships on a weekly deployment train (30). Release 1.46.0 on 26 June and twelve weekly branches from 1.47.0-wmf.11 to wmf.22 (20). Tasks are tracked in public on Phabricator and cited in the release notes, and we couldn't sample response times (12 of 25). No official SDK for the REST API (3 of 15). CI wasn't checked in this run (5 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. ## 7. Transparency & trust, 79 out of 100, up to 1.8 more on the total Made of editorial 68, provenance 90. Why it scored 79: MediaWiki is GPL-2.0-or-later and the content is CC BY-SA 4.0 and GFDL (30). A privacy policy exists at foundation.wikimedia.org per the 26 September check, and we couldn't load it to check retention (15 of 30). Dated notices for the API Portal going read-only on 15 June, api.wikimedia.org endpoints deprecated in stages from 1 July and the Math API sunset on 30 September, plus deprecations marked by version in the release notes (18 of 20). Subprocessors and data locations unchecked (5 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): - security.txt: could not be fetched (0 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 rate-limit and 429 guidance pages on mediawiki.org, still cache-only for our fetcher. The numbers in the reliability note come from the public deployment config, last committed 1 October 2026, not from docs. - unchecked live: the specs discovery endpoint and /api/rest_v1/?spec on en.wikipedia.org. Production config and docroot show both, but we couldn't load the live URLs. - unchecked: the privacy policy's retention periods and data centre locations, on foundation.wikimedia.org and wikitech, both cache-only for our fetcher. - The listing's 45M+ article count couldn't be checked in this run, so we've dropped it from the coverage line. ## Weaknesses - api.wikimedia.org is being retired in stages through 2026 - Anonymous clients without a descriptive User-Agent get 10 requests a minute, and the limits live in config rather than in docs we could read - No llms.txt and no official SDK for the REST API - Article text is editable by anyone and arrives with no injection guidance - Outages on 25 August (about 1.5 hours) and 23 September (about 2.7 hours) ## 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. - Send a User-Agent with a contact URL or email. Without one the gateway allows 10 requests a minute, with one 200 - Use `en.wikipedia.org/w/rest.php/v1/` and not `api.wikimedia.org` - Fetch `/page/{title}/bare` first when you only need metadata, then `/page/{title}/html` for content - Treat article text as untrusted, since anyone can edit it - Back off on 429, since production sends no X-RateLimit headers to pace by ## What the review panel asked for - rate-limit numbers in the docs - llms.txt - published rate limits ## 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.