# Fix list: Duffel Flights and Stays API From Anchor Terminal's listing at https://www.anchorterminal.com/tools/duffel, the October 2026 research run, assessed 1 October 2026. Grade B, 66.9 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 Duffel Flights and Stays API: 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, 40 out of 100, up to 7.5 more on the total Why it scored 40: No x402, MPP or L402 (0). Per-unit prices published without login, $3.00 per order, 1 per cent for Managed Content, $2.00 per paid ancillary, 2 per cent FX, $0.005 per search above 1,500 to 1 (20). Sign-up is free and test mode needs no card or contract, per the 30 September check (20). A person signs up in the dashboard to get tokens (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, 60 out of 100, up to 7 more on the total Why it scored 60: Bearer access tokens from the dashboard, separate test and live tokens, revocable. No scopes found (20). Test mode is the only reduced-privilege mode; holds let an order wait for payment, but there's no read-only token we could find (8 of 20). Responses are structured airline and hotel data, no free text from the open web (10). Every response has an `x-request-id`, and orders and payments are visible in the dashboard. No per-call log export documented (7 of 15). PCI DSS v4 Level 1, yearly penetration tests, quarterly ASV scans, a vulnerability disclosure programme with a GPG key, and a trust centre at trust.duffel.com. No security.txt (404 per the 30 September check) or bug bounty (15 of 20). 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. Schema & documentation, 59 out of 100, up to 6.7 more on the total Why it scored 59: No public OpenAPI document found (0). duffel.com/llms.txt and duffel.com/docs/llms.txt both return 404 (0). The API reference states what each endpoint does and the guides cover when to use offer requests, partial offer requests and holds (16 of 20). Typed fields with enums such as `cabin_class` and passenger `type`, required fields marked (13 of 15). An error object with `type`, `code`, `title`, `message` and `documentation_url`, seven error types and per-status guidance for order creation (15). `Duffel-Version` header, a written versioning policy and a public 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. ## 4. Reliability, 70 out of 100, up to 6 more on the total Why it scored 70: Statuspage at duffelstatus.com with per-supplier components and history (20). Twelve incidents since 7 July. Six are marked major by Duffel, among them a Lufthansa Group, Qantas and Aegean search outage on 28 September (4 hours 48 minutes), Air France KLM card payments down on 17 August (5 hours 24 minutes), a Stays partial outage on 3 September and a dashboard failure on 14 September (1 hour 34 minutes). Each was scoped to one supplier or product and none took the core flights API down, so we score the one-major tier rather than several majors, a departure from the checklist (10). Limit published, 60 requests per 60-second window, with `ratelimit-*` headers (15). 429 tells you to retry after `ratelimit-reset`, and order creation has explicit guidance (a 202 means processing, don't retry; use a 130-second timeout) (15). No SLA on the pricing page for either tier (0). 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. Agent ergonomics, 80 out of 100, up to 3.3 more on the total Why it scored 80: `return_offers=false`, `supplier_timeout` and `max_connections` size an offer request, and the offer list can be fetched separately (18 of 25). Cursor pagination with `limit` on list endpoints (20). Typed error codes with a `documentation_url` on every error (20). Retry rules for order creation are documented, but we found no idempotency key on order creation in the docs we read (12 of 20). Few required fields for a search. One maintained SDK (TypeScript); the Python SDK is archived (10 of 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. ## 6. Transparency & trust, 77 out of 100, up to 2 more on the total Made of editorial 63, provenance 90. Why it scored 77: Closed service with a public services agreement under the law of England and Wales (15). Privacy policy names Duffel Technology Limited (company 11188295) as controller, updated 17 June 2026, with a DPA addendum referenced from the services agreement. Retention is "as long as reasonably necessary" with no periods (18 of 30). A versioning policy keeps the previous version 6 months after a new one and promises email notice before breaking changes; v1 ended on 23 January 2025 (20). The security page says Duffel runs entirely on Google Cloud. Transfers are described in general terms and we didn't find a subprocessor list (10 of 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): - security.txt: not found (0 of 10) ## 7. Maintenance & community, 92 out of 100, up to 0.7 more on the total Why it scored 92: Changelog entry on 2026-09-28 (tax and fee breakdowns on each offer and order) and SDK 4.30.0 on 2026-09-18 (30). SDK releases 4.29.0, 4.29.1 and 4.30.0 on 11, 14 and 18 September (20). Outside pull requests (#1192, #1196) merged in September, a public changelog and email support (20 of 25). The TypeScript SDK is current; the Python SDK is archived (12 of 15). CI on every push and pull request, Dependabot and a September lockfile refresh for two advisories (10). 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. - Whether order creation accepts an idempotency key; the docs we read give retry rules but no key - Whether access tokens can be scoped or made read-only - Duffel's subprocessor list, which may sit in the trust centre we didn't open ## Weaknesses - No OpenAPI document, llms.txt or official MCP server - Default rate limit of 60 requests a minute, which Duffel says can change without notice - Searches above 1,500 per order cost $0.005 each and can get you capped under the services agreement - Python SDK archived; TypeScript is the only maintained SDK - No SLA published for pay-as-you-go or enterprise ## 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 `Duffel-Version: v2` on every request or you get a version error - Fetch the single offer again right before creating the order, since an offer goes stale within minutes - Set a 130-second timeout on order creation, and treat a 202 as processing rather than retrying, or you risk a duplicate - Read `ratelimit-remaining` before fanning out searches; the default is 60 requests per 60 seconds - Count offer requests. Above 1,500 per order the surcharge starts and the agreement lets Duffel cap you ## What the review panel asked for - Machine-payable test access - Warn before the ratio bites - Publish a Stays rate example ## 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.