Anrok

# Fix list: Anrok

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/anrok, the October 2026 research run, assessed 8 October 2026. Grade C, 61.6 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 Anrok: 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, 15 out of 100, up to 10.6 more on the total

Why it scored 15: No x402, MPP or L402 (0). Starter is public at $100 per market a month, Custom is quoted by sales, and high-volume Starter sellers may pay additional fees set out in the contract (10). No free tier or card-free trial found. Every seller account has a sandbox, and pre-sale sandbox access is arranged through sales, which we scored as partial (5). An account starts with "Talk to us" or a demo request, and keys are created in the web app (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, 59 out of 100, up to 7.2 more on the total

Why it scored 59: Two credential models. The API key is one revocable Bearer key per seller with a note and an enabled switch, with no scopes found (20 on its own). The MCP server uses OAuth 2.1 with PKCE, dynamic client registration, six scopes and a revocation endpoint, and access follows the user's Anrok role (30 on its own). Scored between the two (25). MCP scopes split transactionsRead from transactionsWrite, and Anrok says voids, refunds and filing approval never reach the model. The API key has no read-only mode (14). Returns customer names and help centre text written by others, with no injection guidance found (3). The MCP docs say all write operations are audited and the OAuth credential appears on the API keys page. No audit log for API calls was found in the help centre (7). SOC 1 Type II, SOC 2 Type II and an annual third-party assessment. No security.txt, disclosure policy or bug bounty found (10).

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. Agent ergonomics, 70 out of 100, up to 4.9 more on the total

Why it scored 70: List endpoints take a limit, default 25 and maximum 100, with no field selection. The MCP server documents 31 tools, and a connection sees only the tools its approved scopes allow (15). Cursor paging plus filters by customer, status, modified time, jurisdiction and filing. List transactions and filings are premium and enabled on request (18). Typed error codes with a documented resolution each, though 429 and authentication failures are plain text (18). createOrUpdate upserts by id and returns a version, and removeTransactions is safe to repeat. No idempotency keys, and we couldn't read the MCP annotations. The docs label each tool Read or Write (13). Every call is a POST with a JSON body and a createOrUpdate needs few fields. No official SDK found on npm, PyPI or GitHub (6).

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. Schema & documentation, 71 out of 100, up to 4.7 more on the total

Why it scored 71: OpenAPI 3.1 at apidocs.anrok.com/openapi.yaml, 26 operations and 41 schemas (25). www.anrok.com/llms.txt exists but links marketing pages, and apidocs.anrok.com has no llms.txt or Markdown docs. The docs are static HTML that reads without JavaScript (3). Operation descriptions state purpose, the invoice lifecycle says which transaction call to use at each step, and premium endpoints are marked. The MCP tools page gives one line a tool. We couldn't read the live tool definitions without an account (15). 45 enums and 68 required lists, with few length or range constraints (11). Request and response examples, cURL samples, typed error codes on 409 responses and a FAQ table of resolutions. 429 and a missing key answer in plain text (14). The path carries v1 and the spec says 1.1, with no public API changelog found (3).

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.

## 5. Maintenance & community, 47 out of 100, up to 4.6 more on the total

Why it scored 47: Closed service with no release tags. The API docs and spec were last rebuilt on 6 October 2026 per their Last-Modified header, and the help centre article on the MCP server was created on 28 September and updated on 7 October 2026 (30). No dated API changelog. Help centre articles updated in August and September 2026 show activity, and the newest product announcement is 7 July 2026 (6). No public changelog. Support by email and a help centre (6). No official SDK, and the MCP server isn't in the official registry (0). No packages to assess. The docs site is current (5).

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.

## 6. Transparency & trust, 54 out of 100, up to 4 more on the total

Made of editorial 38, provenance 70.

Why it scored 54: Closed service with clear terms, last updated July 2026, naming Anrok, Inc. of Delaware and California law (15). The terms say Customer Data isn't used to train AI models, while Aggregated Data may train Anrok's models. The privacy notice, effective May 2024, gives no retention periods, and the DPA is available on request only (13). No deprecation policy found. One schema in the spec is marked deprecated with no date (3). The privacy notice says the service is hosted in the United States and the security page names Google Cloud. No subprocessor list found (7).

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: anrok.com, registered 2019-01-06 (7 years) (11 of 15)
- Terms of service: read, states 6 of the 7 things a reader expects, and has 2 clauses that cost points (5.1 of 10)
- Privacy policy: read, states 7 of the 8 things a reader expects (9.3 of 10)
- Changelog: not found (0 of 10)
- security.txt: not found (0 of 10)

## 7. Reliability, 88 out of 100, up to 2.4 more on the total

Why it scored 88: Graded on the public API, with the MCP server noted where it differs. Statuspage at anrok-status.com with two components, Web Application and Billing Integrations and API (20). No incident in the 90 days to 8 October 2026. The latest was 10 March 2026, elevated latency and 429 responses for about four hours (30). The spec and the FAQ give a default limit of 10 requests a second per seller account (15). 429 carries an integer Retry-After and the FAQ recommends retries at a fixed interval or with backoff. createOrUpdate overwrites by transaction id and removeTransactions is described as idempotent, but there are no idempotency keys (13). No SLA found on the public pages (0). The API is version 1.1 with no beta label, and the MCP docs carry none either (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.

## 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 live MCP tool definitions, their input schemas and annotations, which need an Anrok login. The count of 31 comes from the tools reference page
- unchecked: the Anrok CLI named on the headless tax page. We found no documentation or package for it
- Whether an SLA exists in order forms. None is published
- Whether the app keeps an audit log of API key use. The help centre article on keys doesn't mention one
- Whether Starter can be bought without a sales call. The pricing page button reads "Talk to us"
- Which plans include MCP. The pricing page lists it under Starter, and the help centre says to ask the account team
- lastRelease is the Last-Modified date of the docs and spec, since Anrok publishes no API changelog

## Weaknesses

- No self-serve signup. Starter is $100 per market a month through "Talk to us", and Custom is quoted by sales
- API keys have a note and an enabled switch only. No scopes, read-only mode or expiry found in the help centre article
- No API changelog, deprecation policy or official SDK found. The spec is version 1.1 with no dated history
- List transactions, filings, tax ID validation, certificate requests and street-level international addresses are premium and enabled on request
- No security.txt, disclosure policy or public subprocessor list. The DPA is available on request only

## 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 every call as a POST with a JSON body to https://api.anrok.com, including reads such as /v1/seller/products/list
- Call createEphemeral for a quote and createOrUpdate only when the invoice is final. createOrUpdate saves the transaction for filing
- Reuse the same transaction id on a retry. createOrUpdate overwrites by id, so a repeat doesn't create a second record
- Expect transactionFrozenForFiling once a transaction is on a filed return. Negate it with createNegation and write a new id with a suffix such as :s2
- On 429 wait the number of seconds in Retry-After. The default limit is 10 requests a second per seller account

## 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.