# Fix list: Middesk From Anchor Terminal's listing at https://www.anchorterminal.com/tools/middesk, the October 2026 research run, assessed 8 October 2026. Grade C, 59 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 Middesk: 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, 10 out of 100, up to 11.3 more on the total Why it scored 10: No x402, MPP or L402 (0). No public price. Fees are set in an order form and middesk.com/pricing returns 404 (0). A sandbox with mock data comes with an account, and the vendor's plugin README says trial accounts issue test keys. We couldn't establish how a trial is opened or whether it needs a card, so half credit (10). Access starts with a demo request and a dashboard sign-in (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, 56 out of 100, up to 7.7 more on the total Why it scored 56: Revocable API keys split by environment, sub-account keys managed through the API, and OAuth 2.0 with `read_only` and `read_write` scopes and a revoke endpoint. OAuth access tokens don't expire, there's no consent screen, and an API key carries full account access (24). A `read_only` token is limited to GET, the sandbox is a separate host and key, and an IP allowlist applies. The MCP server advertises only `read_write`, and no approval step for billable writes was found in the API (12). Responses carry text from websites, adverse media and an LLM-written analyst summary, and no guidance on prompt injection was found (3). Actions record who changed a business and what changed, and a timeline endpoint lists events. No log of API calls for the operator was found (7). SOC 2 Type II audited annually and a Vanta trust centre. 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, 56 out of 100, up to 7.2 more on the total Why it scored 56: List endpoints return a simplified business item, and a few reads take `include`, but the full Business object is large (the Markdown reference page for create business is 1.4 MB) and field selection through GraphQL is early access. The MCP server has eleven tools per the vendor's README (14). `page` and `per_page` with `has_more` and `total_count`, and filters on signals, mail and sub-accounts. The spec shows no filters on the business list (15). Errors are a JSON array of messages with an optional `parameter`, with no machine-readable codes (12). No idempotency key header. `external_id` is an idempotency key for sub-accounts only, lien continuations return the one in flight, and MCP annotations couldn't be read (8). Only `name` and `addresses` are required to create a business, but no official SDK was found in any language (7). 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. Reliability, 73 out of 100, up to 5.4 more on the total Why it scored 73: Graded on the REST API with the hosted-service lines. Statuspage at status.middesk.com with six components and incident history (20). From 10 July to 8 October 2026 it lists 13 incidents. Most are state registry sources (California, Texas, Nevada, Tennessee) running slow or offline. Two are marked major, about three hours of delayed outbound email on 28 July that blocked Middesk Agent sign-in, and ten minutes of intermittent sandbox availability on 1 September. Neither took the production API down for an hour, so we read the record as minor incidents only (20). 20 requests a second per customer is published (15). 429 is documented, with a `Retry-After` header only for sandbox business creation, and safe retries are documented for sub-account creation and lien continuations but no general idempotency key was found (8). No SLA in the public terms (0). REST v1 is generally available (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. ## 5. Transparency & trust, 61 out of 100, up to 3.4 more on the total Made of editorial 43, provenance 78. Why it scored 61: Closed service with public terms last updated 21 August 2026, and MIT plugins (15). The terms bar training generative AI models on customer data and limit ML training to de-identified data. The privacy policy of 2 November 2025 keeps data as long as necessary with no periods, and no public DPA was found (14). The API change policy promises advance notice and an opt-out for breaking changes without a notice period, the terms give 60 days' notice before data is withdrawn, and the changelog labels breaking changes (10). Socure is named for identity checks. No subprocessor list or data location was found outside the trust centre, which we couldn't read (4). 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: middesk.com, registered 2018-11-20 (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 4 of the 8 things a reader expects (7 of 10) - security.txt: not found (0 of 10) ## 6. Maintenance & community, 64 out of 100, up to 3.2 more on the total Why it scored 64: Latest changelog entry 5 October 2026 (30). Nine dated entries between 20 July and 5 October (20). Closed service with a weekly changelog, a help centre and support in standard hours under the terms. No public issue tracker or community channel was found (9). No official SDK, and the MCP server isn't in the official registry. The middesk/plugins repository has four commits since 16 September 2026 (3). No published packages, and the plugins repository has no CI (2). 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. Schema & documentation, 82 out of 100, up to 2.9 more on the total Why it scored 82: Public OpenAPI 3.1 spec with 62 paths and 87 operations (25). llms.txt, and every docs page is served as Markdown by adding .md (10). 41 of 87 operations carry a description, and the guides explain orders, lifecycle and review tasks, with little on when not to call an endpoint (12). 148 enums and required fields in the spec, though `GET /v1/businesses` declares no query parameters and several 422 bodies are typed as any (11). Example responses and a status code table are on the docs pages. The spec itself holds no examples, and errors are a message with an optional parameter name (9). Versioned as v1 with a dated changelog (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. ## 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 MCP server's tool definitions, input schemas and annotations. tools/list needs a token, so the count of eleven comes from the vendor's plugin README - unchecked: the trust centre at trust.middesk.com, which only renders with JavaScript. The SOC 2 report, any subprocessor list and any DPA may sit there - unchecked: how a trial account is opened. The plugin README says trial accounts exist and issue test keys only. The site shows only a demo request, and app.middesk.com/signup is a JavaScript shell - Whether an SLA exists in order forms. None is in the public terms - Whether customers had notice before the 17 August 2026 lien `packet_number` change. The policy says affected customers are told in advance, and we could see only the changelog entry - Whether production 429 responses carry `Retry-After`. The changelog documents it for sandbox business creation only - Whether `unique_external_id` makes `POST /v1/businesses` safe to retry. The spec types it as a string with no description - Per-check prices, which are set in an order form ## Weaknesses - No public price. Fees are set in an order form, and every docs page ends with a prompt to contact sales - No official SDK found on npm, PyPI or RubyGems, and none in the vendor's GitHub organisation - The hosted MCP server runs against production only, so a sandbox key fails on every call - Creating a business places billable orders, and an empty `orders` array doesn't prevent that - OAuth access tokens don't expire, and no security.txt, disclosure policy or bug bounty was found ## 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. - Match the key to the host. `mk_test` keys work only at https://api-sandbox.middesk.com/v1 and `mk_live` keys only at https://api.middesk.com/v1 - Name the `orders` on `POST /v1/businesses`. Omitting them places a verification order plus every package the account runs automatically, all billed - Send `address_line1` and `address_line2`. The API ignores `address_line_1` without an error - A 201 means the business was created, not verified. Wait for the `business.updated` webhook or poll until `status` leaves `pending` - Stay under 20 requests a second per account, and in sandbox wait the seconds in `Retry-After` after a 429 on business creation ## 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.