# Fix list: Convoy From Anchor Terminal's listing at https://www.anchorterminal.com/tools/convoy, the October 2026 research run, assessed 8 October 2026. Grade B, 62.2 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 Convoy: 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, 30 out of 100, up to 8.8 more on the total Why it scored 30: Graded on Convoy Cloud. No x402, MPP or L402 (0). Flat plan prices are public, Pro at $99 and Premium at $499 a month, with no per-message price (10). A 14-day trial without a card, limited to 100 events a day (20). A person has to sign up in a browser to get a Cloud key (0). The self-hosted Community edition is free and `convoy bootstrap --with-api-key` prints a key without a signup, which this score doesn't count because the hosted service is what we graded. 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, 53 out of 100, up to 8.2 more on the total Why it scored 53: Bearer API keys. A project key is limited to one project, a personal key to its user's organisations, and both can be regenerated or revoked. No OAuth for API clients, and we found no per-action scopes on a key (22). Roles include a read-only project viewer, listed as paid for self-hosted instances and on Cloud Premium. Whether an API key can be issued read-only wasn't established, and deletes need no confirmation (8). Incoming webhook payloads are third-party content. The docs cover signature verification and SSRF but give no guidance for agents reading payloads (5). Every delivery attempt is logged with request and response. No audit log of API or dashboard actions was found (7). No security.txt and no SECURITY.md. GitHub private vulnerability reporting is on, one advisory was published in July 2026, and the repository runs CodeQL, OSV and Trivy workflows. The pricing page lists SOC 2, and the trust centre couldn't be read (11). 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, 69 out of 100, up to 5 more on the total Why it scored 69: List sizes are set with `perPage` (default 50), with no field selection or summaries, and event bodies come back whole (12). Cursor pagination in both directions, with filters by date range, endpoint, source and idempotency key (20). Errors are `{"status": false, "message": ...}` with conventional HTTP codes. Messages are specific, but only four codes are documented and there are no machine-readable error codes (10). `idempotency_key` on event creation plus replay and retry endpoints. Endpoint and subscription creates have no idempotency key, and the onboarding guide says to look up by `ownerId` first (15). An event needs only an endpoint id, an event type and data. SDKs for JavaScript, Python, Go and Ruby, with convoy.js last published on 13 November 2023 (12). 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, 78 out of 100, up to 3.6 more on the total Why it scored 78: OpenAPI 3.0 spec in the public repository, 68 operations on 49 paths with 153 schemas, and the API reference is generated from it (25). llms.txt, with every documentation page also served as Markdown (10). Every operation has a description, but most are one line such as "This endpoint retries an event delivery" and none says when not to use it (10). 32 enums and typed request models, though `models.CreateEvent` marks no field as required (10). 58 examples. Every operation lists 400, 401 and 404, while the errors page documents four codes and one sample body, and 429 isn't in the spec (8). Dated API versions with `X-Convoy-Version`, a compatibility matrix and a per-release CHANGELOG.md. The spec's own version field reads 26.3.5 against release 26.8.0 (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. ## 5. Transparency & trust, 67 out of 100, up to 2.9 more on the total Made of editorial 65, provenance 69. Why it scored 67: The gateway's source is public under the Elastic Licence 2.0, which isn't an OSI licence and bars offering it as a hosted service. The OpenAPI spec still names the MPL 2.0 (20). A privacy notice dated 1 June 2023 that still cites Privacy Shield, a DPA with deletion within 60 days of termination and breach notice without undue delay, and a Cloud subscription policy with dated suspension and deletion steps. The terms name no legal entity (15). The Cloud upgrade policy promises at least 180 days' notice of major upgrades and deprecations, and breaking API changes are listed by version (20). US and EU regions are named and self-hosted telemetry is documented with an opt-out (`CONVOY_ANALYTICS_ENABLED=false`). The sub-processor list at trust.getconvoy.io didn't render for us (10). 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: getconvoy.io, registered 2021-09-06 (5 years) (11 of 15) - Endpoint on the vendor's domain: is not on getconvoy.io (0 of 15) - Privacy policy: read, states 5 of the 8 things a reader expects (7.8 of 10) - security.txt: not found (0 of 10) ## 6. Maintenance & community, 80 out of 100, up to 1.8 more on the total Why it scored 80: v26.8.0 was tagged on 27 September 2026, 11 days before this check (30). 22 version tags between 27 June and 27 September 2026 (20). The issues page shows 34 open issues, the most recently updated on 14 September 2026 and many untouched since 2023. Support runs through a community Slack and GitHub (12). convoy-python 0.2.0 was published on 21 July 2026 and a workflow regenerates SDKs from the spec, but convoy.js on npm dates from 13 November 2023 (10). The repository has integration, end-to-end, lint, CodeQL and nightly OSV workflows. We couldn't read their current pass state (8). 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. Reliability, 92 out of 100, up to 1.6 more on the total Why it scored 92: Graded as a hosted service, Convoy Cloud. status.getconvoy.io on incident.io lists the website and the US and EU data and control planes, with data since July 2024 (20). It shows 100 per cent uptime on every component from July to October 2026 and no incidents, though the day-by-day calendar didn't load for us (30). Cloud Pro is limited to 25 events a second, and the docs give defaults of 1,000 API requests and 1,000 ingested events a second for self-hosted instances (15). The docs say a breach returns 429 and document idempotency keys on events, but give no Retry-After or backoff guidance for API clients, although the server code sets Retry-After (10). The pricing page lists a 99.99 per cent uptime SLA on Pro and 99.999 on Premium. No SLA document was found (7). The API 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. ## Deductions Each comes off the total. A fixed and documented problem counts for less at the next check. - 2026-07-24. Advisory GHSA-p5vg-v7mj-f6q4, rated High. Before v26.6.8 any caller authorised on one project could read another project's source record by id, including message broker credentials in plaintext. Patched in 26.6.8 and published by the maintainers, so the deduction is reduced to 4 (https://github.com/frain-dev/convoy/security/advisories/GHSA-p5vg-v7mj-f6q4). - 2026-08-04. Until v26.7.0 the events list returned `metadata` on dynamic events, which carried the endpoint secret and custom auth headers in plaintext. The field was removed across every API version and the change is documented, so the deduction is 2 (https://www.getconvoy.io/docs/api-reference/versioning). ## 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 trust centre at trust.getconvoy.io and its sub-processor list render only with JavaScript, so SOC 2 status and sub-processors rest on the pricing page's claim and the DPA's link - unchecked: the day-by-day incident calendar on status.getconvoy.io didn't load. The 90-day record rests on the page's 100 per cent uptime figures and its empty incident feed - unchecked: the GitHub API refused us for its rate limit. Stars (2,877) come from the repository page, and the pass state of CI on main wasn't read - Whether an API key can be issued with a read-only role on Convoy Cloud wasn't established from the docs - No SLA document was found behind the uptime percentages on the pricing page - The Cloud signup page wasn't exercised, so the no-card trial rests on the pricing page - First release date not established from a 200-commit clone ## Weaknesses - Advisory GHSA-p5vg-v7mj-f6q4 (24 July 2026, High) let a caller on one project read another project's source and broker credentials before v26.6.8 - Until v26.7.0 (4 August 2026) the events list returned endpoint secrets and custom auth headers in plaintext for dynamic events - No security.txt and no SECURITY.md. The trust centre at trust.getconvoy.io renders only with JavaScript, so we couldn't read it - The errors page documents four HTTP codes and one sample body. 429 and Retry-After aren't in the API reference - Cloud needs a browser signup, and the 14-day trial allows 100 events a day, one project and one user ## 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. - Use the regional base URL, https://us.getconvoy.cloud/api/v1 or https://eu.getconvoy.cloud/api/v1. A project key works only under /projects/{projectID}/ - Create an endpoint, then a subscription, then the event. An event sent to an endpoint with no subscription isn't dispatched - Send `idempotency_key` on every event. A repeated key creates the event but no delivery, and the key stays reserved until retention deletes the event - Create projects with a personal API key and the `orgID` query parameter. The project key in that response is shown once - Before retrying an endpoint or subscription create, list endpoints by `ownerId`. Idempotency keys cover event ingestion only ## 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.