# Fix list: Kestra From Anchor Terminal's listing at https://www.anchorterminal.com/tools/kestra, the October 2026 research run, assessed 8 October 2026. Grade B, 63.6 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 Kestra: 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. Security & auth, 41 out of 100, up to 10.3 more on the total Why it scored 41: The open-source edition has one basic-auth username and password with full access, and no tokens or scopes (10). No read-only mode or approval step in open source. RBAC, service accounts and human approval tasks are Enterprise. MCP servers are private by default (3). Allow and deny lists restrict outbound HTTP from tasks. No guidance on untrusted content reaching an agent was found (5). Audit logs are Enterprise only. Open source keeps execution history and logs, and MCP calls are labelled `system.from:mcp` (7). A valid security.txt, a SECURITY.md with a two-business-day acknowledgement, 33 advisories published by the maintainers, and SOC 2 Type 2 for the company and the managed service. No bug bounty found (16). 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. ## 2. Payments & pricing, 50 out of 100, up to 6.3 more on the total Why it scored 50: Read with the self-hosted rule. No x402, MPP or L402 (0). The open-source edition is free with unlimited flows and executions, stated on the pricing page. The paid options have no public price, Enterprise through sales and Cloud by access request, so half marks (10). Free to run with no card (20). An agent can start the Docker image and set the credential in configuration with no signup (20). 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. Agent ergonomics, 73 out of 100, up to 4.4 more on the total Why it scored 73: List and search calls take `page` and `size`, and execution search returns a light execution type. No field selection (15). Paging, sorting and operator filters on search endpoints (20). Errors are problem+json, though most operations document only the generic 401, 403 and 500 (13). No idempotency key. The documented pattern passes `system.correlationId` as a label and needs a duplicate check written into the flow. Flows exposed as MCP tools can set read-only, destructive and idempotent hints (10). Executing a flow needs only namespace and flow id, and there are generated SDKs for Python, JavaScript, Java and Go (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. ## 4. Schema & documentation, 82 out of 100, up to 2.9 more on the total Why it scored 82: OpenAPI 3.0.1 file for the open-source API at kestra.io/kestra.yml, 216 operations on 194 paths, versioned 2.0.5 (25). `llms.txt`, `llms-full.txt` and every docs page as Markdown by adding `.md` (10). 213 of 216 operations have a summary, most of them one line saying what the call does and none saying when to choose it (11). 49 enum schemas and bounds such as `size` 1 to 1,000, but execution inputs are an untyped multipart array and filters are strings in a nested query syntax (10). 401, 403 and 500 are documented as problem+json on 213 operations, 409 on 23 and 400 on 22, and the API guide has curl examples (11). The path carries `v1`, the spec carries the release number, and release notes and migration guides are public (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, 69 out of 100, up to 2.7 more on the total Made of editorial 86, provenance 51. Why it scored 69: Apache-2.0 for the server and the SDK repository (30). For self-hosted use the security page says flows, logs and secrets stay in the owner's database and storage, which agrees with the usage-reporting page. The Cloud privacy policy and terms cover only Cloud, and the website policy only the website, so no document covers the usage reports as personal data (20). A written release policy gives each long-term support line a support end date, and removals come with migration guides. No notice period for API changes was found (16). Usage reporting is on by default, documented field by field with a link to the source, and both streams can be switched off (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): - Domain age: kestra.io, registered 2019-12-18 (6 years) (11 of 15) - Endpoint on the vendor's domain: is not on kestra.io (0 of 15) - Terms of service: not found (0 of 10) - Privacy policy: not found (0 of 10) - Status page: not found (0 of 10) ## 6. Reliability, 89 out of 100, up to 2.2 more on the total Why it scored 89: Graded as software the owner runs, on the open-source edition 2.0.5. Official Docker image `kestra/kestra` and a stated runtime of Java 25 (20). Public CI with unit and end-to-end tests. The latest `main-build` run on the `develop` branch on 8 October 2026 passed, after one failure two hours earlier (22). 840 open issues and pull requests. The nine issues opened on 7 and 8 October were all labelled and had one to twelve comments each (17). Major, feature and patch versions follow a written release policy, and each version's breaking changes have a migration guide (15). Version 2.0, with 1.0 released on 9 September 2025 (15). 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. ## 7. Maintenance & community, 93 out of 100, up to 0.6 more on the total Why it scored 93: v2.0.5 and v1.3.42 released on 5 October 2026 (30). 32 releases between 15 July and 5 October 2026 (20). New issues are labelled and answered within a day in the sample we read, against 840 open issues and pull requests (20). SDKs for four languages are generated from the spec, with `kestrapy` 2.0.1 published on 11 September 2026 (15). CI, CodeQL and scheduled end-to-end tests run on the repository. Two points off because the docs say `kestrapy` 2.0.1 imports a package it does not declare (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. ## Deductions Each comes off the total. A fixed and documented problem counts for less at the next check. - Five advisories rated critical were published on the repository in six months. GHSA-365w-2m69-mp9x (CVE-2026-34612, remote code execution through SQL injection, 30 March 2026), GHSA-5vc5-wxxq-3fjx and GHSA-2q47-568g-9h4f (CVE-2026-49869 and CVE-2026-53576, unauthenticated remote code execution through authentication filter bypass, 3 June 2026), and GHSA-rjhm-qm6w-m7x9 and GHSA-j5cv-8rw9-vv2p (unauthenticated remote code execution and authentication bypass, 29 September 2026). All are fixed, in 1.3.38 and 1.0.60 at the latest, and the maintainers published each one, so the deduction is reduced (https://github.com/kestra-io/kestra/security/advisories). - Two further high advisories affected the default open-source setup. GHSA-hrr4-xg8h-5p6f, an unauthenticated gRPC control plane on port 50051, fixed in 2.0.3 and published 29 September 2026, and GHSA-94pv-f379-3gp3, a revoked administrator credential that stayed valid, fixed in 1.3.41 and published 6 October 2026 (https://github.com/kestra-io/kestra/security/advisories). ## 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: status.kestra.io did not answer our requests, so we could not tell whether Kestra Cloud has a status page - unchecked: the trust centre on app.drata.com (sub-processor list, SOC 2 report, policies) was not read - Kestra Cloud's rates for task runs and runner time are not published, and Enterprise is priced through sales - The lead said the docs give four language SDKs. That holds, though the repository README names only Python and JavaScript - Whether the open-source `default` MCP server can be limited to fewer flows per caller, given the single credential - kestra.io/terms returns 404. The only terms found are the Kestra Cloud Terms of Service, which exclude the open-source project ## Weaknesses - The open-source edition has one basic-auth username and password. API tokens, service accounts, RBAC and audit logs need Enterprise or Cloud - Five critical advisories from 30 March to 29 September 2026, four of them unauthenticated remote code execution or authentication bypass, all fixed - Kestra Cloud is request-access with no public price, and Enterprise is sold by annual contract through sales - No rate limits and no idempotency key on the API. The correlation ID pattern needs a guard written into the flow - The docs say `kestrapy` 2.0.1 imports `regex` without declaring it, so a plain install fails on import ## 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. - Run 2.0.5 or 1.3.42 or later. Earlier builds carry unauthenticated remote code execution advisories fixed in 1.3.38 and 2.0.3 - Set `kestra.server.basic-auth` in the config file before first start. Without it the setup page is open to anyone who reaches the port - Put the tenant in the path. Open-source instances use `main`, as in `/api/v1/main/executions/{namespace}/{id}` - Send flow inputs as multipart form fields, and add `wait=true` to get the finished execution in the same call - Keep port 8081 off any shared network. The management endpoints there skip the API's basic auth ## 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.