# Fix list: Lingo.dev From Anchor Terminal's listing at https://www.anchorterminal.com/tools/lingo-dev, the October 2026 research run, assessed 8 October 2026. Grade B, 62.3 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 Lingo.dev: 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, 20 out of 100, up to 10 more on the total Why it scored 20: No x402, MPP or L402. The API's 402 status reports a credit limit and carries no payment instructions (0). Unit prices are public without a login, $2 per million output tokens plus the model's cost with no markup, $0.50 per engine a month and $0.01 per AI review run (20). Sandbox has a $0 base price, but every translation is charged against prepaid credits and no free allowance or trial was found on the pricing page or in the docs (0). A person signs up in a browser. The CLI signs in with an emailed code and the MCP server with browser OAuth (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. Schema & documentation, 54 out of 100, up to 7.5 more on the total Why it scored 54: No OpenAPI file was found. `https://api.lingo.dev/openapi.json` returns 404 and the docs link none. The open-source JavaScript SDK carries typed request schemas, and the MCP tool definitions can't be read without signing in (5 of 25). `llms.txt` returns one title line and docs pages have no Markdown version (0 of 10). The reference says what each endpoint is for and when to pick the sync call or async jobs (16 of 20). Parameters have types and limits, such as 1 to 100 target locales, 100 locked-key patterns and five job statuses, while `data` is a free-form object by design (11 of 15). Request and response examples are given in up to five languages, and one page lists seven status codes. Error bodies hold only a `message` string (12 of 15). A weekly changelog runs from 15 March to 1 October 2026. No API version in the path or a header was found (10 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. ## 3. Security & auth, 62 out of 100, up to 6.7 more on the total Why it scored 62: The MCP server uses OAuth with PKCE and dynamic client registration and publishes no scopes. API keys are stored hashed, shown once and revoked by deletion. Personal keys carry their creator's full access, and keys with their own role and engine scope need Enterprise. No documented option puts a key in a URL (22 of 30). RBAC with 20 permissions exists on Enterprise only. On other plans every member is an Owner. The MCP server can create and revoke API keys and invite members with admin access by default, and no confirmation step is documented (8 of 20). Responses are translations of the caller's own content, while engine provisioning reads links the caller supplies. No injection guidance was found (7 of 15). Request logs with model, tokens and full input and output are on every plan. The append-only audit log, which records the acting user or key and the IP address, is Enterprise only (12 of 15). security.txt gives a contact and a disclosure policy with a safe-harbour promise and a maximum reward of a $50 gift card. The site states SOC 2 Type II, with the report available on Enterprise. No bug bounty or public advisories were found (13 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. ## 4. Reliability, 71 out of 100, up to 5.8 more on the total Why it scored 71: Hosted reading of the checklist, for the REST API and the hosted MCP server. status.lingo.dev runs on OpenStatus with WEB and API monitors and an events history (20). The history lists two partial API outages in the last 90 days. One opened on 18 July 2026 and was marked resolved 19 hours later, the other ran from 22:09 to 23:59 UTC on 22 September 2026. Both were partial and both passed an hour, so we score the record as one major outage (10). Daily token quotas are published per plan, 100,000 on Sandbox and 5 million on Production, with no per-second request limit found (12 of 15). The errors page says to retry 500 with exponential backoff and not to retry 402 or 429, and async job creation takes an `idempotencyKey`. No `Retry-After` header is documented (12 of 15). The pricing page lists a 99.99% uptime SLA for Enterprise without publishing the document, and the terms say the company strives for 99.9% without a guarantee (7 of 10). Version 1.0 shipped on 15 March 2026 and nothing on the API pages is marked beta (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. Agent ergonomics, 75 out of 100, up to 4.1 more on the total Why it scored 75: API reading of the checklist. Sync responses return only the translated data, the model and a usage block, and job lists return summaries without output. The MCP server covers team, role, key and audit administration alongside translation, with no documented toolsets or read-only subset, and its tool count is unpublished (15 of 25). Job lists page by cursor with `limit` from 20 to 100 and filters for engine and status (20). Errors are keyed on seven status codes with retry advice for each group, but the body has no machine-readable code (15 of 20). Async job creation is idempotent with `idempotencyKey`. The sync call has no such key, and MCP annotations could not be read (12 of 20). A sync call needs three fields and falls back to the organisation's default engine. JavaScript and Python SDKs exist, though the Python one last shipped on 7 March 2026 and the docs site has no SDK section (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. ## 6. Transparency & trust, 71 out of 100, up to 2.5 more on the total Made of editorial 51, provenance 91. Why it scored 71: The platform is closed under published terms. The legacy CLI, the JavaScript SDK, the compiler and the Python SDK are Apache-2.0, and the current `@lingo.dev/cli` package states no licence (18 of 30). The terms and the privacy policy agree that user content is not used to train AI models and that AI providers keep content for up to 30 days by default. The privacy policy gives translation data 30 days by default while the pricing page gives 7 days on Sandbox. The Data Processing Agreement the policy refers to is not published (17 of 30). The terms promise advance notice and migration instructions for breaking API changes without stating a period, and the older CLI's docs are kept under a legacy section (8 of 20). The terms name Supabase in Frankfurt and Cloudflare. The sub-processor page the privacy policy points to, lingo.dev/security, returned 404, and the LLM providers are not listed as sub-processors on any page we read (8 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): - Domain age: lingo.dev, registered 2021-05-03 (5 years) (11 of 15) - Terms of service: read, states 7 of the 7 things a reader expects, and has 2 clauses that cost points (6 of 10) - Privacy policy: read, states 6 of the 8 things a reader expects (8.5 of 10) ## 7. Maintenance & community, 87 out of 100, up to 1.1 more on the total Why it scored 87: `lingo.dev` 0.138.10 shipped on 7 October 2026 and `@lingo.dev/cli` 1.17.1 on 6 October (30). `@lingo.dev/cli` had eight releases since 10 July 2026 and the changelog has an entry every week (20). The GitHub repository has 5,402 stars and 17 open issues, each with at least one comment and most of them feature requests. Two compiler bug reports from March 2026 remain open. The weekly changelog is public and the site links a Discord community (17 of 25). The official MCP registry lists `dev.lingo/main` 1.0.0 at https://mcp.lingo.dev/main, a different path from the `/account` endpoint in the docs. The JavaScript SDK last shipped on 24 July 2026 and the Python SDK on 7 March 2026 (12 of 15). The repository runs pull-request checks and a release workflow and bumped dependencies on 7 October. `@lingo.dev/cli` states no licence, repository or engines field on npm (8 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 definitions, tool count and annotations. The server answers 401 without an OAuth token and the docs list capabilities, not tools. - unchecked: whether a new account receives starting credits and whether signup needs a card. The pricing page and docs state neither and the dashboard is behind a login. - unchecked: the Data Processing Agreement and the sub-processor list. The privacy policy points to lingo.dev/security, which returned 404, and no published DPA was found. - unchecked: the SOC 2 Type II report. The site footer states it and the pricing page lists it for Enterprise. - The terms list using any robot, crawler, scraper or other automated means to access the services as prohibited, while the product is sold as an API, CLI and MCP server. Recorded as a fact with no deduction. - The terms' fees section speaks of custom quotes, while the pricing page publishes unit prices. We used the pricing page. - The MCP registry entry `dev.lingo/main` points at https://mcp.lingo.dev/main, and the platform docs give https://mcp.lingo.dev/account. We did not establish how the two endpoints differ. - The lead's docs link, https://lingo.dev/en/mcp, redirects to https://lingo.dev/en/docs/mcp. The lead named the vendor Lingo.dev, and the legal entity is Replexica, Inc. - The legacy CLI sends usage events to PostHog unless `DO_NOT_TRACK=1` is set, per its source, and `@lingo.dev/cli` depends on `posthog-node`. We did not find a docs page that discloses this. ## Weaknesses - No public OpenAPI file was found, `llms.txt` holds one title line and docs pages have no Markdown version - Organisation keys with their own role and engine scope, RBAC and audit logs are Enterprise only. On other plans every member is an Owner - The status page lists two partial API outages in 90 days, on 18 July 2026 (closed after 19 hours) and 22 September 2026 (1 hour 50 minutes) - Errors carry a status code and a free-text `message`, with no machine-readable code and no `Retry-After` documented - The privacy policy points to lingo.dev/security for sub-processors, and that page returned 404 on 8 October 2026 ## 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 `X-API-Key` to https://api.lingo.dev and keep the key server-side. On non-Enterprise plans a personal key reaches every engine its creator can - Use `POST /process/localize` for one language pair and `POST /jobs/localization` with an `idempotencyKey` for several. Pipeline stages run on async jobs only - Treat 402 (credit limit) and 429 (daily token quota) as limits to report, not to retry. Retry 500 with backoff - Read `usage.cost` on each sync response to track spend. Sandbox stops at 100,000 tokens a day - Connect MCP clients to https://mcp.lingo.dev/account. Its tools can create and revoke API keys and invite members, so ask before those calls ## 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.