# Fix list: Secoda From Anchor Terminal's listing at https://www.anchorterminal.com/tools/secoda, the October 2026 research run, assessed 9 October 2026. Grade E, 44.5 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 Secoda: 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, 0 out of 100, up to 12.5 more on the total Why it scored 0: Hosted service sold by contract, so the hosted rubric applies. No x402, MPP or L402 (0). The pricing page names Core, Premium and Enterprise with no prices, and every plan button links to a sales form (0). No free tier or trial was found on the site. The API docs still say to sign up for free at app.secoda.co, which we did not confirm (0). An agent needs a user in a customer workspace to generate an API key (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, 40 out of 100, up to 12 more on the total Why it scored 40: Hosted service, read on the hosted lines and graded on the REST API, with the local MCP server as a second surface. No status page is linked from the home page, the pricing page or the documentation index, and we did not try guessed addresses (0 of 20). With no page there is no readable incident history (5 of 30). Writes are limited to 30 PUT, PATCH or POST calls a minute on entity endpoints. No limit for GET requests was found (10 of 15). A 429 is documented with advice to retry after a delay or back off. No `Retry-After` header or idempotency key was found (6 of 15). The Service Level Agreement attached to the subscription agreement commits to 99.9 per cent monthly availability with service credits (10). The REST API carries no beta label. The MCP server is at version 0.1.0, and the Secoda AI Terms still describe a private beta (9 of 10). 0 + 5 + 10 + 6 + 10 + 9 = 40. 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, 44 out of 100, up to 9.1 more on the total Why it scored 44: Graded on the REST API. The catalogue response carries more than 40 fields for each resource, and no field selection or page size parameter was found. The MCP server is small, six tools with short definitions (12 of 25). Lists take `page` and return `links`, `meta` and `total_pages`, with JSON `filter` and `sort` and a filter generator in the app's settings. Results stop at 10,000 resources (15 of 20). Errors are HTTP status codes with one-line descriptions, plus a troubleshooting note on trailing slashes (6 of 20). No idempotency keys. Updates use PATCH, bulk update takes 100 entities, and the MCP tools set no `readOnlyHint` or `destructiveHint` (4 of 20). Few required parameters. No official API client libraries were found, and the Secoda SDK is for building custom integrations that run inside Secoda (7 of 15). 12 + 15 + 6 + 4 + 7 = 44. 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, 54 out of 100, up to 8.1 more on the total Why it scored 54: A workspace API key sent as a Bearer header, with the same access as the user who created it. No scopes or expiry were found, and we did not confirm how a key is revoked. No secret travels in a query string (18 of 30). A key follows its creator's role and team membership, and `run_sql` needs an admin toggle and query access to the integration. No read-only key and no confirmation step for deletes were found (8 of 20). The API and MCP tools return descriptions, documents and query results written by workspace users. No prompt-injection guidance was found (3 of 15). An audit log filterable by user, date, type and resource, an audit log endpoint in the API, and SIEM logging on Enterprise (12 of 15). A disclosure policy with a reply within 5 business days and safe harbour, a bug bounty by email with a reward for findings of CVSS 4 or higher, and SOC 2 stated with documentation on request. No `security.txt` and no public advisories (13 of 20). 18 + 8 + 3 + 12 + 13 = 54. 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, 59 out of 100, up to 6.7 more on the total Why it scored 59: Each API reference page embeds OpenAPI 3.0 definitions at version 8.22.66, with request and response schemas. The full description behind the ReDoc page the reference links to, on api.secoda.co, answered 403 to our reader, so we read the fragments on six of 23 reference pages (20 of 25). `llms.txt` with 371 lines and a Markdown copy of every page (10). Operations carry one or two sentences of purpose, deprecated list endpoints name their replacement with a filter example, and List Resources states its 10,000 limit. Lineage parameters have no descriptions, and the six MCP tools have one-line descriptions with a server prompt on choosing between the two search tools (10 of 20). Response fields are typed and described, but `filter` and `sort` are URL-encoded JSON passed as strings, and few enums were seen (7 of 15). An examples page with five workflows and authentication samples in curl, JavaScript and Python. Errors are status codes with a one-line description and no body format (7 of 15). Paths carry `/api/v1` and the definitions carry a version, but the public changelog stops at 28 October 2025 and has no API entries (5 of 15). 20 + 10 + 10 + 7 + 7 + 5 = 59. 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, 58 out of 100, up to 3.7 more on the total Made of editorial 56, provenance 59. Why it scored 58: Closed service with a published Master Subscription Agreement. The MCP server is MIT (15 of 30). A privacy policy, a Data Processing Agreement with 10 days' notice of subprocessor changes and breach notice without undue delay, return of customer data within 30 days of termination, and AI terms that rule out training on customer data without permission. Some statements disagree. The AI security page says OpenAI and Anthropic process in the US while also naming Google, the Data Processing Agreement's security appendix refers to audio and video streams, and no legal document mentions Atlassian (17 of 30). Deprecated endpoints are flagged in the API definitions with a replacement and no date. The acquisition post promises technical plans and timelines in advance, and none were found on the pages read (6 of 20). 25 subprocessors with locations, and three AWS hosting regions (18 of 20). 15 + 17 + 6 + 18 = 56. 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: secoda.co, no registry record we could read (0 of 15) - Terms of service: read, states 5 of the 7 things a reader expects, and has 1 clause that costs points (6.3 of 10) - Privacy policy: read, states 5 of the 8 things a reader expects (7.8 of 10) - Status page: not found (0 of 10) - security.txt: not found (0 of 10) ## 7. Maintenance & community, 60 out of 100, up to 3.5 more on the total Why it scored 60: The docker-compose and Terraform repositories for self-hosted Secoda were tagged v2026.3.1 on 6 October 2026 (30). Five self-hosted versions were tagged between 4 August and 6 October 2026 (20). Closed service, scored on the 15-point scale. The public changelog stops at 28 October 2025, and support is by email and a Slack community. The MCP repository has one open issue (5 of 15). No official API client libraries and no entry in the official MCP registry. The MCP repository's last commit is 26 January 2026 (3 of 15). The MCP repository has no CI, no tests and no tags, and its dependencies are unpinned (2 of 10). 30 + 20 + 5 + 3 + 2 = 60. 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 full OpenAPI description. The ReDoc page the reference links to at api.secoda.co/api/schema/redoc/ answered 403 and we did not retry. - unchecked: 17 of the 23 API reference sections, so the total number of operations is not established. The six sections read hold 31 operations. - unchecked: whether a status page exists. None is linked from the home page, the pricing page or the documentation index, and we did not try guessed addresses. - unchecked: whether app.secoda.co still accepts a new workspace without a contract. The site shows no sign-up button, while the API docs say to sign up for free. - unchecked: how an API key is revoked, whether it expires, and whether users who are not admins can create one. - unchecked: a rate limit for GET requests. - unchecked: Atlassian's plans and dates for Secoda's API and infrastructure. The acquisition post promises them and we read no Atlassian page. - unchecked: www.secoda.co/llms.txt, which redirects to robots.secoda.co, a host we did not fetch. - unchecked: the registration date of secoda.co. The RDAP lookup we used has no service for .co domains. - The lead said no MCP server was seen in the docs. One is documented (https://docs.secoda.co/features/ai-assistant/secoda-mcp-server), and its page lists eight tools where the source registers six. - The pricing page lists API access under the Core plan, and its FAQ says the API is on Business and Enterprise plans, names that no longer appear in the plan table. - The Terms of Use forbid bots that crawl or scrape Secoda, and the subscription agreement forbids disclosing benchmark results. Both matter before any probe is run. ## Weaknesses - No price is published. The three plans (Core, Premium, Enterprise) link to a sales form, and no trial or sign-up button was found on the site - No status page is linked from the home page, the pricing page or the documentation index - The public changelog stops at 28 October 2025 and the self-hosted changelog at 29 October 2025, while self-hosted versions were tagged up to 6 October 2026 - An API key has the same access as the user who created it, with no scopes. No idempotency keys and no error body format are documented - The MCP page lists eight tools, and the published server registers six, with one-line descriptions and no read-only or destructive annotations - The Terms of Use (23 December 2021) forbid bots that crawl or scrape Secoda, and the subscription agreement forbids disclosing benchmark results. Both matter before any probe is run ## 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 by workspace region. `app.secoda.co` uses `api.secoda.co`, `eu.secoda.co` uses `eapi.secoda.co` and `apac.secoda.co` uses `aapi.secoda.co` - Keep writes under 30 PUT, PATCH or POST calls a minute on entity endpoints, and back off on a 429 - List with `/api/v1/resource/catalog` and a URL-encoded JSON `filter`. The per-type list endpoints for tables are marked deprecated, and results stop at 10,000 resources - Send at most 100 entities to `/api/v1/resource/all/bulk_update/` in one request - Add a trailing slash when an endpoint does not behave as expected, as the docs advise - Treat an API key as the creating user. `run_sql` in the MCP server works only when an admin has enabled Run SQL and the user has query access to the integration ## 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.