Sanity

# Fix list: Sanity

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/sanity, the October 2026 research run, assessed 7 October 2026. Grade BB, 73.7 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 Sanity: 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, 40 out of 100, up to 7.5 more on the total

Why it scored 40: No x402, MPP or L402 in the documentation or on the pricing page (0). Plans and unit prices are public. Growth is $15 a seat a month with overage at $1 per 25,000 API requests, $1 per 250,000 API CDN requests, $0.30 a GB of bandwidth and $0.50 a GB of assets (20). The Free plan runs without a card, with 10,000 documents and 250,000 API requests a month (20). A person has to create the account in a browser. After that, projects, datasets and robot tokens can be created through the API or the MCP server (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, 67 out of 100, up to 5.8 more on the total

Why it scored 67: The MCP server uses OAuth with PKCE, dynamic client registration and a revocation endpoint, but lists one scope, `global`. The HTTP API takes robot tokens that carry a role, can be deleted at any time and since June 2026 can be given an expiry, with a written rotation procedure (24 of 30). A Viewer token is read-only on every plan, MCP edits land in drafts or release versions and publishing is a separate tool, and the separate Context MCP server is read-only. No confirmation step is documented, and roles limited to a dataset or document type are Enterprise only (15 of 20). No prompt-injection guidance found in the MCP documentation. Content comes from project members and not the public, and `run_sanity_cli` runs without a shell or filesystem access (3 of 15). Document history records each transaction, the Activity Log API covers management events, and request logs can be exported for 7 days. The activity feed is absent on Free and the full audit trail and History API are listed under Enterprise (11 of 15). SOC 2 Type 2, a vulnerability disclosure programme and a security.txt without an Expires field. The bug bounty pilot has closed and no ISO 27001 certificate of Sanity's own was found (14 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. Reliability, 77 out of 100, up to 4.6 more on the total

Why it scored 77: Graded on the hosted HTTP API and the hosted MCP server. Statuspage at www.sanity-status.com with 25 components, including api.sanity.io, apicdn.sanity.io and MCP Server, and a full incident history (20). Between 9 July and 7 October 2026 it lists three incidents marked major. Studio connectivity on api.sanity.io on 22 July stayed open for 6 hours 20 minutes with two recurrences, API and API CDN errors on 14 August were resolved in 46 minutes, and Content Agent errors on 23 September lasted 59 minutes. Minor incidents include mutation errors for some projects on 15 September. We read that as one major outage of the core API (10). Rate limits are published with numbers, 500 requests a second per IP, 25 mutations and 25 uploads a second per IP, 500 concurrent queries and 100 concurrent mutations per dataset (15). 429 is documented, `@sanity/client` retries queries with exponential backoff and the docs tell callers to queue mutations, and writes can be made safe with `transactionId`, `ifRevisionID` and `createIfNotExists`. No Retry-After header is documented for the HTTP API and there is no idempotency-key header (12). The SLA at sanity.io/legal/sla commits to 99.9 per cent on Enterprise E1 and 99.95 per cent on E2 (10). The HTTP API is generally available and the MCP server has been since v2.6.0 on 11 December 2025 (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.

## 4. Agent ergonomics, 74 out of 100, up to 4.2 more on the total

Why it scored 74: GROQ projections and slices let a caller choose the fields and the number of results, and MCP query responses are limited to 64 KiB. The MCP server documents 53 tools, which scores 5 on the MCP line with no toolsets or read-only subset documented, so we settled on 15 of 25 across both surfaces. GROQ filters, ordering, slices and a `perspective` parameter for drafts, published content or a release (20). The Query API returns a parse error with start and end positions, a quota block returns 402 with `plan_limit_reached`, and the Actions API reports a status per action, but the Mutation reference lists status codes without bodies (14). `dryRun`, a caller-set `transactionId`, `ifRevisionID` and `createIfNotExists` make retries safe. The v2.14.0 changelog says read-only and destructive annotations were added to MCP tools, which we couldn't read without a token (15). Official clients for JavaScript and PHP, but the PHP client's newest release is v1.5.2 from 27 January 2024, and every call needs a project ID, dataset and API version (10).

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. Schema & documentation, 87 out of 100, up to 2.1 more on the total

Why it scored 87: 26 OpenAPI specs covering 225 operations are public at www.sanity.io/docs/api/openapi and through `sanity openapi get` (25). llms.txt, llms-full.txt and a Markdown copy of every documentation page (10). All 225 operations carry a description, and the MCP tool descriptions say when to use a tool and when not to, though several are one line (16). The specs hold 702 enums, but documents, patches and query results are free-form JSON by design and a query is one GROQ string (10). 1,272 example keys across the specs. 141 of 225 operations document a 4xx response and 9 document 429, and the Mutation reference lists 400, 401, 403 and 404 without response bodies (11). Dated API versions in the URL and a public changelog with an entry per release (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, 87 out of 100, up to 1.1 more on the total

Made of editorial 79, provenance 95.

Why it scored 87: Sanity Studio, the CLI and `@sanity/client` are MIT. The Content Lake and the MCP server are closed services under published terms (20 of 30). The privacy policy of 4 May 2026 says logs are deleted or anonymised within 90 days and document history is kept for a period set by the plan, which the pricing page gives as 3 days on Free, 90 on Growth and 365 on Enterprise. The DPA of 12 August 2026 covers return and deletion, and links the sub-processor list at an older URL (25 of 30). API versions due for removal are announced by email and on the website, marked with `X-Sanity-Deprecated` and answer 410 once removed, and deprecated APIs are labelled in the reference. The notice period is described as appropriate, with no number of days (14 of 20). The sub-processor list of 19 January 2026 names 11 companies with processing locations, with Google Cloud in Belgium as the primary host (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 4 of the 7 things a reader expects, and has 1 clause that costs points (5.4 of 10)
- Privacy policy: read, states 7 of the 8 things a reader expects (9.3 of 10)

## 7. Maintenance & community, 89 out of 100, up to 1 more on the total

Why it scored 89: MCP server v2.40.0 on 2 October 2026 and Sanity Studio v6.18.0 on 6 October 2026 (30). The MCP registry lists 30 versions of the server published between 15 July and 2 October 2026, each with a changelog entry or a patch note (20). The open-source Studio repository has 6,352 stars and 284 open issues and pull requests. Two of the three newest open issues had comments when we looked, a small sample, and there is a Discord channel for the MCP server (17 of 25). io.sanity.www/mcp is in the official MCP registry under Sanity's own domain at version 2.40.0, and `@sanity/client` is at 8.9.0 (15). Packages are current and require Node 22.12 or later. We didn't read their CI (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.

## 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: MCP tool input schemas and the read-only and destructive annotations, which need an authenticated session. The count of 53 tools comes from the documentation page
- unchecked: whether the HTTP API sends a Retry-After header on 429. The documentation mentions it only in MCP tool descriptions
- unchecked: CI status of the sanity-io/sanity and sanity-io/client repositories
- The pricing page gives the Growth plan 25,000 documents and the technical limits page gives 50,000. We used the pricing page
- The 22 July 2026 incident is titled as a Studio connectivity problem but is filed against api.sanity.io. We counted it as a major outage of the core API, and the status page doesn't say how many API calls failed
- PyPI download figures were not read, pypistats.org answered 429. Sanity has no official Python client in the documentation we read
- No ISO 27001 certificate of Sanity's own was found on the security page, which lists ISO certifications for Google Cloud
- lastRelease is the MCP server's v2.40.0 on 2 October 2026. Sanity Studio v6.18.0 followed on 6 October

## Weaknesses

- Schema validation rules run only in Sanity Studio. The HTTP mutation API accepts a document without checking them
- Custom roles scoped to a dataset or document type are an Enterprise feature. Robot tokens on other plans take a built-in role across the project
- The MCP OAuth server lists one scope, `global`, and the server documents 53 tools with no toolset or read-only mode
- A status incident on 22 July 2026, marked major on api.sanity.io, stayed open for 6 hours 20 minutes with two recurrences
- The vulnerability disclosure page says the bug bounty pilot has closed and no rewards are paid

## 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.

- Pin a static dated version in every URL, such as `v2025-02-19`. Omitting `apiVersion` in `@sanity/client` falls back to `v1`.
- Validate documents against the schema yourself before an HTTP write, or run `sanity documents validate` afterwards. The Content Lake does not enforce schema rules.
- Back off on 429 for mutations yourself. `@sanity/client` retries queries five times but never retries mutations. The limit is 25 mutations a second per IP.
- Over MCP, call `create_version` before `patch_documents` when editing inside a release, then patch the returned version ID with the same `releaseId`.
- Use GROQ projections and slices to size results. MCP query responses are limited to 64 KiB, and a blocked Free project answers 402 with `plan_limit_reached`.

## 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.