# Fix list: Groundmail API + MCP From Anchor Terminal's listing at https://www.anchorterminal.com/tools/groundmail, the October 2026 research run, assessed 17 October 2026. Grade C, 57.6 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 Groundmail API + MCP: 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. Reliability, 38 out of 100, up to 12.4 more on the total Why it scored 38: Read with the hosted lines, for the API and hosted MCP an agent calls. No status page found. status.groundmail.dev doesn't resolve and groundmail.dev/status answers 404 (0). No readable incident history, and the service launched on 16 October 2026 (5). Sending quotas are published with numbers (3,000 a month and 100 a day on Free, and a new-account ramp-up from 500 to 50,000 a day) and reported in `ratelimit-*` headers, but the short-window limit behind `rate_limited` has no published number (8 of 15). `Retry-After` on 429, a backoff recipe of up to five tries, and `Idempotency-Key` on POST /emails and /events kept 24 hours, with the 504 outcome-unknown case spelt out (15). No SLA, and the terms disclaim uninterrupted availability (0). Not labelled beta or preview, apart from the provider importer (10). 38. 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. ## 2. Payments & pricing, 35 out of 100, up to 8.1 more on the total Why it scored 35: No x402, MPP or L402 in the docs, the OpenAPI document or the pricing page (0). Plan prices, volume tiers and overage per 1,000 published without a login (20). The Free plan needs a card on file before mail reaches anyone but the account email. Without a card only sandbox and self-tests work (5 of 20). The docs say an agent with an AgentID inbox can open its own account, which we didn't try, but sending to other recipients still needs a person's card (10 of 20). 35. 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. ## 3. Security & auth, 56 out of 100, up to 7.7 more on the total Why it scored 56: Scoped API keys with Full, Send only, Read only and custom sets, send keys limited to named domains, and OAuth with PKCE S256, resource indicators, one-hour tokens and refresh-token reuse detection. Bearer tokens go in the header only. An OAuth request that names no scope gets `full` (28 of 30). Read-only and send-only presets for keys and MCP connections, and the sandbox limits sends to the account email until a domain and card are on file. Confirmation before a send is left to the client (16 of 20). The MCP returns subjects and bounce reasons, and the API returns full inbound messages from /received-emails. No prompt-injection guidance found (4 of 15). Connected CLIs and agents are listed with their scopes and can be revoked, keys show `last_used_at`, and GET /emails logs each send. No request or audit log is documented (8 of 15). No security.txt, disclosure policy, bug bounty or certification found (0 of 20). 56. 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. Transparency & trust, 48 out of 100, up to 4.6 more on the total Made of editorial 41, provenance 55. Why it scored 48: Closed service with published terms, and the MCP server and SDKs under MIT (20 of 30). The privacy policy lists data categories and names Google Cloud, Stripe, Redis, Sentry, Google Analytics, PostHog, Crisp and HubSpot by role. Message retention is 'according to plan and configuration', and a DPA and subprocessor list are on request only. The terms say they took effect on 8 August 2026, before the domain existed (10 of 30). No deprecation policy. The terms promise notice of material changes by email, and the errors page says existing codes keep their meaning (3 of 20). Data may be processed 'in the United States and other regions', with no list of locations (8 of 20). 41. 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: groundmail.dev, registered 2026-10-16 (under a year) (0 of 15) - Status page: not found (0 of 10) - Changelog: not found (0 of 10) - security.txt: not found (0 of 10) ## 5. Schema & documentation, 77 out of 100, up to 3.7 more on the total Why it scored 77: OpenAPI 3.1 at docs.groundmail.dev/openapi.yaml with 70 operations, each naming its scope in `x-required-scope`, and typed input and output schemas on all 20 MCP tools (25). llms.txt, a 267 KB llms-full.txt and Agent Skills (10). MCP descriptions are one line each, with a few 'do not combine' rules and a detailed import tool, and only 21 of 70 OpenAPI operations carry a description (12 of 20). The OpenAPI document bounds `limit` from 1 to 100 and uses enums, but the MCP `limit` is an unbounded number, `to` takes one address and template `variables` are free-form (11 of 15). An errors page with every code, its status, a fix and examples, 69 of 70 operations listing 4xx responses, and curl and SDK examples (15). The OpenAPI document is versioned 0.12.0 and the MCP repository has semver tags, but there's no public changelog (4 of 15). 77. 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. Agent ergonomics, 79 out of 100, up to 3.4 more on the total Why it scored 79: 20 MCP tools in about 14 KB of definitions, with no toolsets. Scoped connections make unpermitted tools fail with `insufficient_scope` rather than hiding them (15 of 25). `limit` and `cursor` on every list, and a domain filter on received mail, but no status or date filters on GET /emails (14 of 20). Errors carry `code`, `hint`, `docs_url`, `required_scope` and `upgrade_url`, and the MCP passes them through. A few older endpoints return `error` as a string, which the docs flag (19 of 20). Idempotency keys on the API, and readOnly, destructive and idempotent hints on every MCP tool, but `send_email` has no idempotency input (16 of 20). Two required fields to send, and SDKs for Node, Python, Go, PHP and Ruby (15). 79. 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. ## 7. Maintenance & community, 72 out of 100, up to 2.5 more on the total Why it scored 72: groundmail-mcp 0.6.3 and groundmail 0.14.7 published on 17 October 2026 (30). 15 versions of groundmail-mcp in two days, a launch burst rather than a cadence (20). The GitHub API was closed to this session, so issue replies couldn't be read. Both public repositories are one day old and mostly bot mirror commits, there's no public changelog, and support is by email (3 of 25). In the official MCP registry under the GitHub-verified namespace io.github.Groundmail, with current SDKs (15). Tests in the MCP and SDK repositories and a current MCP SDK dependency (^1.30.0), but no CI workflow in the public repositories, whose commits are mirrored from a private one (4 of 10). The project is one day old, so these lines reflect the checklist. 72. 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: GitHub issues, pull requests and stars, because the GitHub API was closed to this session - unchecked: the hosted MCP's tool list, which needs a sign-in. It is assumed to match groundmail-mcp 0.6.3 - unchecked: the PyPI, RubyGems and Packagist pages for the SDKs, which weren't read. The SDK list comes from the docs and the SDK repository - The short-window request rate limit behind `rate_limited` isn't published - Whether the 209.188.111.0/24 pool, registered to SuperSend, carries only transactional mail as Groundmail says, could not be established - Why the terms say they took effect on 8 August 2026 when the domain was registered on 16 October 2026 - The MCP README's mention of schedule and cancel tools was treated as a documentation error, not a deduction ## Weaknesses - Launched on 16 October 2026, with no status page, changelog, SLA or incident history - A card on file is needed to send to anyone but the account email, including on the Free plan - No security.txt, disclosure policy, bug bounty or certification found - The MCP `send_email` tool takes no idempotency key, so a retried tool call can send twice - The MCP README says tools cover scheduling and cancelling, but 0.6.3 has no tool for either ## 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 an `Idempotency-Key` on raw POST /emails and reuse it on retries. On 504, check GET /emails before sending again - Test from `noreply@mail.groundmail.dev` to the account email only. Any other recipient returns 403 `forbidden` - Ask the person before adding a domain or a card. Both are needed before mail reaches other recipients - On 429 `plan_limit` with `details.reason: "warmup"`, wait until `details.resets_at`. Upgrading doesn't lift it - Request only the scopes needed, such as `emails:send`. An OAuth request with no scope gets `full` ## What the review panel asked for - a card-free path to a first outside send - a documented agent account route - a public changelog - a status page - Lift caps on upgrade - State how refused sends count - least-privilege OAuth default - mark inbound mail untrusted ## 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.