# Fix list: Alation From Anchor Terminal's listing at https://www.anchorterminal.com/tools/alation, the October 2026 research run, assessed 8 October 2026. Grade C, 60.3 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 Alation: 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, so the hosted rubric applies. No x402, MPP or L402 (0). No published prices. www.alation.com/pricing redirects to a form for a 30-minute sales call (0). No free tier or trial for a new user. The 2,000 free Aggregated Context API calls are a one-time allowance inside a paying tenant (0). An agent needs a customer's Server Admin to register an OAuth client, or a user to create a token, after a sales contract (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, 61 out of 100, up to 7.8 more on the total Why it scored 61: Hosted service, read on the hosted lines and graded on what an agent uses, the Alation Cloud Service REST APIs and the MCP server. A Statuspage at status.alationcloud.com with production and development components for nine regions (20). The history page lists four incidents between 10 August and 10 September 2026. US-East-1 lost connectivity for 2 hours 26 minutes on 10 September (marked critical), eu-west-1 was disrupted for 65 minutes on 10 August (major), a subset of customers had degraded availability for 56 minutes on 18 August (major), and Sydney was down for 19 minutes on 20 August (critical). Each was regional, and two ran over an hour. July was not on the page we read (5 of 30). The AI API has a two-tier token bucket with no numbers published. REST throttling is off by default and, when enabled, defaults to 20 requests a second for each API family. The Aggregated Context API has a one-time quota of 2,000 calls with enforcement at 2,400 (10 of 15). AI API 429s carry `Retry-After` and three `Ai-RateLimit` headers, REST 429s state the seconds to wait, and the SDK marks 429 and 500 as retryable. No idempotency keys were found (10 of 15). The MSA sets a 99.5 per cent monthly uptime target with service credits of 1 to 5 per cent (10). The Aggregated Context API has been generally available since 2025.1.5. The AI API page calls itself new and under highly active development, and the SDK and local MCP server are at 1.0.0rc3 (6 of 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. Security & auth, 63 out of 100, up to 6.5 more on the total Why it scored 63: OAuth 2.0 client credentials on Alation Cloud Service. A Server Admin registers a client application, which creates a system user with one Alation role, and the admin can set token duration, revoke tokens, rotate the secret and rotate the signing key. An authorisation code flow with refresh tokens exists for user sign-in, without dynamic client registration. API keys are a refresh token (60 days) and an access token (24 hours) sent in a `TOKEN` header, revocable by an admin. Permissions follow roles, not scopes, and the refresh token endpoint takes a username and password (24 of 30). The APIs by Roles table maps each API to the roles that may call it, AI SQL execution allows reads only, and tools can be disabled. No confirmation step for writes was found (13 of 20). Tools return descriptions, documents and query text written by catalogue users, and we found no prompt-injection guidance (3 of 15). A Logging API, an API usage log, and SDK tool events sent to the customer's instance with tool name, parameters, duration and status. The source says parameters are not redacted (10 of 15). ISO 27001, 27701, 27017 and 27018, SOC 2 Type 2 with HIPAA, two penetration tests a year and a security@alation.com address. No security.txt, no bug bounty, and advisories need a Community login (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. Agent ergonomics, 71 out of 100, up to 4.7 more on the total Why it scored 71: The local MCP server registers 16 tools and enables 8 by default, with `--enabled-tools` and `--disabled-tools` lists, and the README says the remote server lists only the tools the current user may call. Descriptions are long. On the Aggregated Context API a signature selects object types and fields, and field caps hold a table to its top 50 columns (22 of 25). REST paging uses `limit` and `skip` with an `X-Next-Page` header and a 1,000-object cap, and the AI API uses `limit` and `offset` up to 1,000. Parameter names differ between API families (17 of 20). Six-digit error codes with a published table, and the SDK returns `reason`, `resolution_hint`, `is_retryable` and `help_links` (17 of 20). No idempotency keys in the docs and no `readOnlyHint` or `destructiveHint` in the MCP source. Bulk writes run as jobs with a status endpoint (6 of 20). The catalogue search agent needs one parameter. The official SDKs (`alation-ai-agent-sdk`, `allie-sdk`, `alation-data-quality-sdk`) are all Python (9 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. ## 5. Schema & documentation, 78 out of 100, up to 3.6 more on the total Why it scored 78: Every API has an OpenAPI 3.0 definition, printed in the Markdown copy of each reference page and served by each instance under `/openapi/`. We found no single public file to download. The local MCP server's tools take typed parameters through FastMCP (22 of 25). No llms.txt on developer.alation.com, www.alation.com or docs.alation.com, but the developer portal serves a Markdown copy of each page at the same address with `.md` (6 of 10). MCP tool descriptions name use cases, counter-examples and the tool to call instead, and the API reference states purpose and availability for each operation (16 of 20). The `signature` parameter is a free-form JSON object, search `filters` is a JSON string, and the agent tools take one natural-language `message` (9 of 15). A table of six-digit error codes, request and response examples and recipes. The portal warns that its code generator builds a wrong URL for `filters` (13 of 15). Reference pages are versioned from v2024.1 to v2026.9.0-1 and paths carry v1 or v2, but the API release notes page we read holds only the 2026.7.2.0 entry (12 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. Maintenance & community, 71 out of 100, up to 2.5 more on the total Why it scored 71: The status page shows the 2026.9.1.0 upgrade reaching US instances on 3 October 2026 (30). Upgrades to 2026.7.2, 2026.8.0, 2026.9.0 and 2026.9.1 appear between 1 August and 3 October (20). Closed service, scored on the 15-point scale. API release notes and a support policy with response targets by severity. On GitHub the agent SDK has two open issues, one from 10 September 2026 asking when 1.0 will ship and a bug from 7 November 2025 (8 of 15). `alation-ai-agent-sdk` and `alation-ai-agent-mcp` were last published on 14 January 2026 as 1.0.0rc3, `allie-sdk` 2.2.3 on 25 May 2026 and `alation-data-quality-sdk` 1.0.7 on 14 May 2026. No entry in the official MCP registry (6 of 15). The SDK repository runs Renovate and test workflows with actions pinned to commit SHAs, and its last commit is dated 4 August 2026 (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. ## 7. Transparency & trust, 76 out of 100, up to 2.1 more on the total Made of editorial 65, provenance 86. Why it scored 76: Closed service under the published Master Cloud Software Licence and Services Agreement of 10 April 2026. The agent SDK and MCP server are Apache-2.0 (15 of 30). The MSA deletes customer data within 30 days of termination, the Data Privacy Addendum promises breach notice without undue delay, and the AI Acceptable Use Policy bars training on customer data while keeping anonymised usage data and derived learnings. The Aggregated Context API guide says content is not shared with external model providers and also says the question and signature go to a language model routed through Amazon Bedrock (22 of 30). An end-of-support matrix for versions and connectors. The SDK gave three months' notice for removing `alation_context` in February 2026, and the tool was still in the source in August. No API deprecation policy with periods was found (10 of 20). A sub-processor list with locations, 30 days' notice of additions, and a table of model routing regions (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 5 of the 7 things a reader expects, and has 1 clause that costs points (6.3 of 10) - Privacy policy: read, states 7 of the 8 things a reader expects (9.3 of 10) - security.txt: not found (0 of 10) ## 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 Alation Cloud Service documentation that moved to the Help Centre on 18 August 2026. help.alation.com showed our reader an error page, so the remote MCP server's set-up, its tool list and any audit log pages went unread. - unchecked: the remote MCP server's tool schemas and annotations at `https:///ai/mcp`, which need a customer instance. - unchecked: status history before August 2026. The history page we read listed August, September and October, and robots.txt on status.alationcloud.com disallows the JSON API. - unchecked: numeric limits for the AI API's two token-bucket tiers. The docs name the tiers and headers without numbers. - unchecked: security advisories since March 2024, which sit behind an Alation Community login, and the SOC 2 and ISO documents, which go through a sales representative. - unchecked: comment counts and reply dates on the two open GitHub issues. - The API release notes page we read holds one entry, 2026.7.2.0 (updated 30 July 2026), while the portal's version menu runs to v2026.9.0-1. Later API notes were not found. - The agent SDK README said in December 2025 that `alation_context` would be removed in February 2026. It was still in the source on 4 August 2026. - firstReleased is left empty. The agent SDK first reached PyPI on 2 May 2025, and we did not establish the catalogue's own launch date. - popularity.pypiWeekly is `allie-sdk` (2,087). `alation-ai-agent-sdk` had 113 downloads in the week and `alation-ai-agent-mcp` 93. ## Weaknesses - No published price, trial or self-serve sign-up. The pricing address redirects to a sales form - Four incidents between 10 August and 10 September 2026, two of them regional outages longer than an hour - `alation-ai-agent-sdk` and `alation-ai-agent-mcp` were last published on 14 January 2026 as 1.0.0rc3, and a plain `pip install` still resolves to 0.12.0 - No guidance on prompt injection, although tools return catalogue descriptions, documents and query text written by users - No `security.txt` and no bug bounty, and security advisories sit behind a Community login ## 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. - Ask a Server Admin to register an OAuth client application with the lowest role that fits. Only admins can create one, and the secret is shown once - Pin `alation-ai-agent-mcp==1.0.0rc3`. An unpinned install resolves to 0.12.0, which predates the catalogue search agent - Call `catalog_context_search_agent` for catalogue questions. `alation_context` is deprecated and its description tells models not to call it directly - Watch `X-Entitlement-Usage` on Aggregated Context API responses. The 2,000 free calls never reset, and a quota 429 is not retryable - Page with `limit` and `skip` and follow `X-Next-Page`. The BI Source API uses `offset`, and one call returns at most 1,000 objects ## 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.