# Fix list: Contentstack From Anchor Terminal's listing at https://www.anchorterminal.com/tools/contentstack, the October 2026 research run, assessed 7 October 2026. Grade B, 64 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 Contentstack: 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: Read with the hosted rubric. No x402, MPP or L402 (0). Plan prices are public without login ($0, $29 and $299 a month, with included API calls stated). Build mentions pay-as-you-go overages and no unit rate is shown (10). The Free plan needs no card and doesn't expire. We didn't run the sign-up form to confirm it (20). A person signs up in a browser, verifies an email address and creates the token or approves 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. Reliability, 71 out of 100, up to 5.8 more on the total Why it scored 71: Read with the hosted lines and scored on the Content Management API. status.contentstack.com is a Statuspage site with components by cloud, region and service and a full incident history (20). Its feed lists nine incidents with customer impact between 9 July and 7 October 2026, each confined to one or two of the seven regions. The longest on the core APIs was a disruption of the management, delivery and automation APIs on Google Cloud North America on 11 July (74 minutes, marked minor). Webhook delivery on AWS Europe was disrupted for 59 minutes on 28 August (marked critical), and publishing and webhooks on Azure Europe were delayed for 51 minutes on 1 September (marked major). We read that as one major outage with several shorter ones (10). Rate limits have numbers, 10 reads and 10 writes a second per organisation and one bulk request a second (15). The docs say a 429 is returned and name `X-RateLimit-Limit` and `X-RateLimit-Remaining`. No Retry-After header, backoff guidance or idempotency key was found (6 of 15). The Services Description commits to 99.50 or 99.95 per cent monthly availability by plan, with service credits (10). The v3 API is generally available. The MCP server is at version 0.9.0 (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. ## 3. Agent ergonomics, 67 out of 100, up to 5.4 more on the total Why it scored 67: Responses can be sized with `only[BASE][]`, `except[BASE][]` and `limit`. The MCP server's default group is 78 tools and about 66 KB of schema, split from nine other groups by `--groups` (17 of 25). `limit` up to 100, `skip`, `include_count`, sort and a JSON `query` filter (20). Errors use standard status codes with a JSON body, with 412 and 422 explained in the docs. Error bodies aren't in the spec (12 of 20). No idempotency keys or conditional writes were found, and the MCP package contains no readOnlyHint or destructiveHint annotations. Entries keep version numbers, which helps check a write after a lost response (4 of 20). Management SDKs in JavaScript, Python, Java and .NET. Every call needs two headers and the right regional host (14 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. Security & auth, 69 out of 100, up to 5.4 more on the total Why it scored 69: OAuth 2.0 with scopes per module and action (cm.entry:read, cm.entry:write and cm.entry:publish are separate), 60-minute tokens, a refresh grant and revocation by users or organisation admins. Management tokens can be read-only, branch-limited and expiring, and are shown once. PKCE isn't mentioned in the OAuth docs (27 of 30). Read-only tokens and separate publish scopes allow least privilege, and stacks have workflows and publish rules. The MCP server has no read-only switch and no confirmation step, and its default group includes nine delete tools (13 of 20). Entries are written by a stack's own users and can still carry instructions. No prompt-injection guidance was found in the MCP docs (3 of 15). A stack audit log records creates, updates, deletes and publishes by user, with an API endpoint and MCP tools to read it. Its retention period isn't stated (13 of 15). SOC 2 Type II, ISO 27001:2022 and penetration tests twice a year. Reports go to security@contentstack.com with a stated 24-hour response, and SDK security fixes are listed in the changelog with CVE numbers. security.txt returns 404 and no bug bounty was found. We didn't search NVD (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. ## 5. Schema & documentation, 77 out of 100, up to 3.7 more on the total Why it scored 77: An OpenAPI 3.0.0 file for the Content Management API is public, with 138 paths and 221 operations, and every MCP tool has a JSON Schema published at mcp.contentstack.com (25). www.contentstack.com/llms.txt indexes the docs. The full-text export it points to, llms-full.txt, returned 502 on 7 October (8 of 10). Operation descriptions in the spec are long (median 1,189 characters, in HTML) and MCP tool descriptions say what each tool does (median 142 characters in the `cma` group). Few say when not to use a tool (15 of 20). The spec has no enums and no component schemas, and request bodies are examples. The `cma` tool schemas have 369 properties with two enums and 13 untyped objects such as `entry_data`. Version 0.9.0 says it adds enums from the connected stack at run time, which we couldn't see without an account (7 of 15). The spec carries 494 examples but documents only a 200 response for each operation. The docs list status codes in a table (9 of 15). The API version is in the path, and a dated changelog with an RSS feed covers API, SDK and CLI changes. A behaviour change to `limit` shipped inside v3 (13 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, 73 out of 100, up to 2.4 more on the total Made of editorial 59, provenance 87. Why it scored 73: The service is closed under a Master Agreement last updated 17 July 2026 and Terms of Service from August 2022. The spec, SDKs, CLI and MCP package are MIT, though the MCP repository named on npm isn't public (15). The privacy policy of 30 June 2026 says customer content is deleted within 180 days of termination, data processing agreements for the US and for EMEA are published, and the AI addendum of 27 July 2026 says customer data isn't used to train general AI models. The older Terms of Service only say access to content ends on termination (24 of 30). No deprecation policy with a notice period was found. The Services Description lets Contentstack end a service during a term with a pro-rata refund, and the `limit` change was announced on the day it applied (5 of 20). The sub-processor list was updated on 25 May 2026 with purposes and locations. It gives the United States for Azure and Google Cloud although both have European regions in the API docs, and the page states no notice period for changes (15 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 1 clause that costs points (7.1 of 10) - security.txt: not found (0 of 10) ## 7. Maintenance & community, 82 out of 100, up to 1.6 more on the total Why it scored 82: The JavaScript management SDK 1.31.2 was released on 22 September 2026, and the changelog's newest entry is 1 October 2026 (30). Dated entries in the last 90 days include the `limit` change on 11 September, recursive embedded items on 7 September, webhook HMAC signing on 30 July and more than ten SDK and CLI releases (20). Closed service with a public changelog and RSS feed, email support and a Discord community. We didn't test a support channel (10 of 15). Official management SDKs in four languages, all with releases since 27 July 2026. The MCP server isn't in the official MCP registry, which returned no result for Contentstack (15). The JavaScript SDK repository has unit-test, CodeQL and dependency-scan workflows, and its changelog shows regular dependency fixes. We didn't see the CI results (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. ## Deductions Each comes off the total. A fixed and documented problem counts for less at the next check. - 11 September 2026. The `limit` query parameter on delivery and management requests changed behaviour inside v3. `limit=0` used to return every matching record and now returns the default 100, so an integration relying on it gets a truncated result with a 200 status. The changelog entry is dated the same day and gives a migration path, and no earlier notice was found in the changelog feed. Deducted 3, the low end, because it was documented (https://www.contentstack.com/docs/changelog). ## 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 endpoint-level API reference on the docs site, which renders in the browser. We read the introduction and the OpenAPI file - unchecked: https://www.contentstack.com/llms-full.txt returned 502 on 7 October 2026 - unchecked: GitHub star counts, and whether CI passes on the SDK repositories - unchecked: the sign-up form itself. The pricing page says no card is required - unchecked: NVD and GitHub advisories for Contentstack packages - unchecked: whether customers were told by email before the 11 September 2026 `limit` change. Only the public changelog was read - The docs disagree on how many management tokens a stack may hold. The API page says 10 and the token guide says 30 - The run-time enrichment of MCP tool schemas in 0.9.0 needs a connected stack, so we graded the static definitions - The Enterprise card on the pricing page says 99.5 per cent uptime while the Services Description names plans (Start, Grow, Scale, X1, X3, X5) that don't match the pricing page's plan names. Whether Free, Build and Growth carry any commitment wasn't established ## Weaknesses - The MCP server's default group loads 78 tools, including nine deletes, with no readOnlyHint or destructiveHint annotations and no read-only switch - On 11 September 2026 `limit=0` stopped returning every record and now returns the default 100. The changelog entry is dated the same day - The OpenAPI file documents only 200 responses and has no enums or component schemas. Request bodies are shown as examples - No idempotency keys and no Retry-After header were found. The docs name only X-RateLimit-Limit and X-RateLimit-Remaining - No security.txt and no bug bounty were found, and no deprecation policy with a notice period - Nine incidents with customer impact on status.contentstack.com between 9 July and 7 October 2026, each in one or two regions ## 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. - Pick the base URL for the stack's region first. North America on AWS is https://api.contentstack.io, and the other six regions use contentstack.com hosts - Send `api_key` and `authorization` headers on every Content Management API call. Ask for a read-only management token when the task only reads - Page with `limit` (100 at most), `skip` and `include_count=true`. `limit=0` no longer returns everything - Stay under 10 reads and 10 writes a second per organisation, and one bulk request a second. Watch `X-RateLimit-Remaining` and back off on 429 - Start the MCP server with `--groups cma` only, and add `cma-extended` when the task needs audit logs or version history. Publishing and deleting need no confirmation ## 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.