Crowdin

# Fix list: Crowdin

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/crowdin, the October 2026 research run, assessed 8 October 2026. Grade B, 69.6 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 Crowdin: 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, 30 out of 100, up to 8.8 more on the total

Why it scored 30: No x402, MPP or L402 (0). Four plan prices are public in the pricing page's structured data (Free $0, Pro $50, Team $150, Team+ $450 a month). Hosted-word limits and the CDN add-on's unit prices are drawn by script and were not read, and machine and AI translation are paid from prepaid credits with no rate found, so plan-only credit (10). A Free plan exists, with a 14-day trial of Team (20). A person signs up in a browser. The MCP server's dynamic client registration still ends in a person's sign-in (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. Security & auth, 68 out of 100, up to 5.6 more on the total

Why it scored 68: Personal access tokens with an expiry date, scopes at read-only or read-and-write, per-project granular access and revocation. OAuth authorisation code grant with PKCE for public apps. No query-string token option found (30). Read-only scope levels and the MCP `crowdin-read-only` header. No server-side confirmation for destructive calls was found, and the MCP examples include deleting a user (15 of 20). The API returns strings, comments and translations written by contributors. No prompt-injection guidance was found in the developer or help documentation (0 of 15). A security log of account events (logins, password changes, device and IP address) with a read-only `security-log` scope, and AI request logs. No per-call API log found (8 of 15). security.txt valid to 2030, a vulnerability policy pointing to a private HackerOne programme, ISO/IEC 27001 certification and HIPAA agreements claimed on the security page. No SOC 2 report found (15 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. Agent ergonomics, 69 out of 100, up to 5 more on the total

Why it scored 69: The MCP V2 server has 33 tool sets, all loaded by default. We could not list the tools without an account, so the count is read as more than 30 (5), with 10 added back for the tool-set, project-type and wildcard filters and the read-only mode (15 of 25). `limit` up to 500, `offset`, `orderBy` and CroQL filters on the API, and `crowdin://` resources on MCP (20). Errors are JSON with a message and code, or per-field string codes. The 429 has no recovery guidance (15 of 20). No idempotency keys. The MCP read-only header helps, and tool annotations could not be read (6 of 20). Seven official clients and a CLI, though a file upload takes two calls through storage (13 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.

## 4. Reliability, 78 out of 100, up to 4.4 more on the total

Why it scored 78: Hosted reading. The surface graded is the REST API v2, with the hosted MCP server noted where it differs. status.crowdin.com is a Statuspage site with ten components, three of them for the API (20). Its incident feed shows nothing between 10 July and 8 October 2026. The latest entries are a Crowdin Translate downtime report on 7 July and a minor slowdown on 16 March (30). The reference states 20 simultaneous requests an account across all tokens (15). A 429 is returned above that, with no Retry-After header, backoff guidance or idempotency keys found. Long jobs have a documented start, status and download sequence (3 of 15). No SLA found in the terms or the help centre (0). API v2 and MCP V2 are presented as current, with no beta label (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.

## 5. Schema & documentation, 80 out of 100, up to 3.3 more on the total

Why it scored 80: Public OpenAPI 3.0 document for API v2, 374 operations on 218 paths in the file-based reference, with string-based and Enterprise variants (25). `llms.txt` links abridged and complete documentation and one text file per route (10). Operations carry a summary and the required scope, and parameters point to the call that supplies their value, but few say when to use or avoid an endpoint (12 of 20). 518 enums, required fields and numeric bounds. Edits are JSON Patch arrays and several forms are `oneOf` (12 of 15). 2,831 examples, with 400, 401, 403, 404 and 405 responses on most operations (13 of 15). The API is versioned at `/api/v2` and the MCP server at V1 and V2. Changes appear in monthly product posts and client release notes, and no dated API changelog was found (8 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.

## 6. Transparency & trust, 79 out of 100, up to 1.8 more on the total

Made of editorial 65, provenance 93.

Why it scored 79: Closed service with published terms revised on 24 September 2026 and earlier versions kept on GitHub. The CLI and clients are MIT (15). The privacy policy gives retention periods (closed-account data deleted within 60 days, backups up to two more months, billing records 7 years), the terms carry a data processing contract, and both say Crowdin does not train models on client input. For managed AI the commitment about sub-processors is commercially reasonable efforts (26 of 30). MCP V1 and 12 API operations are marked deprecated with no removal dates, and no deprecation policy was found (6 of 20). A dated sub-processor list with countries, and data centres named as the United States and Ireland (18 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 6 of the 7 things a reader expects, and has 3 clauses that cost points (3.1 of 10)

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

Why it scored 82: `crowdin-api-client` 1.30.0 on PyPI on 30 September 2026, `@crowdin/crowdin-api-client` 1.59.0 on 29 September and `@crowdin/cli` 5.3.0 on 24 September (30). Six CLI releases and four JavaScript client releases since 20 August, and monthly product posts for July, August and September (20). Monthly dated product updates and a support contact. The CLI and JavaScript client repositories each show 2 open issues per the GitHub API. We did not read the issue threads (12 of 15). Seven current official clients (15). CI and dependency state were not inspected, so half credit (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.

## 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 MCP server's tool count and tool definitions, including annotations. The endpoint answers 401 without a token and we created no account.
- unchecked: plan limits, the plan comparison table and the CDN add-on's unit prices on crowdin.com/pricing, which are drawn by script.
- unchecked: whether the Free plan includes API and MCP access, and whether it needs a card. Neither is stated in the text we could read.
- unchecked: rates for Crowdin Credits (machine translation and AI).
- unchecked: CI state and issue threads of the client and CLI repositories, and the official MCP registry.
- The category fits loosely. Crowdin is a localisation management platform with machine and AI translation calls inside it, and no formality control was found in the OpenAPI document.
- The terms forbid accessing the service to monitor availability, performance or functionality or for benchmarking without written consent, which matters before any probe is run. Recorded as a fact, no deduction.
- No SLA, deprecation policy, SOC 2 report or dated API changelog was found in the pages read. An Enterprise SLA may exist in order forms.

## Weaknesses

- The terms forbid accessing the service to monitor availability, performance or functionality, or for benchmarking, without Crowdin's prior written consent
- A 429 is returned above 20 simultaneous requests an account, with no Retry-After header or backoff guidance in the reference
- No idempotency keys in the OpenAPI document, and no SLA found in the terms or the help centre
- Plan limits and add-on unit prices on the pricing page are drawn by script. Only four plan prices were readable, from the page's structured data
- MCP V1 and 12 API operations are marked deprecated with no removal date, and no deprecation policy was found

## 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 `Authorization: Bearer <token>` to `https://api.crowdin.com/api/v2`. Crowdin Enterprise uses `https://<organization_domain>.api.crowdin.com/api/v2`
- Use the MCP V2 endpoint `https://mcp.crowdin.com/v2/mcp` and set `crowdin-tool-sets`, because the server loads all 33 tool sets by default
- Set `crowdin-read-only: true` on the MCP connection for exploration. It exposes only list and get tools
- Keep to 20 concurrent requests an account across all tokens, and add your own backoff on 429
- Upload a file to `/storages` first, then pass the `storageId` to the add-file call. Builds, exports and reports are asynchronous and need a status poll
- `/mts/{mtId}/translations` takes at most 100 strings a call and needs the `mt` scope and a configured engine

## 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.