Not an API linter — an agent-readiness check. Paste an OpenAPI spec and get a 0–100 score for how well an LLM agent can discover and call it.
Live scans of well-known public APIs — these are the actual numbers.
Six categories an agent depends on, weighted by how much they affect tool-calling.
Unique operationIds and a clear summary so an agent knows what each tool is and when to call it.
Described, typed parameters — enums where bounded, required flags set — so agents fill them correctly.
Typed success responses with examples an agent can actually parse, not a bare 200 OK.
Documented 4xx errors with schemas, so an agent can recognise and recover from failures.
Declared auth schemes and destructive operations flagged as high-risk.
Pagination, tags, and a tool count that fits the agent's context window.
Every scan lists what's wrong. Unlock to get the fix for each one. A real example:
GET /usersOperation has no operationId — agents have no stable name to call it by.
Add a unique, verb-led operationId (e.g. listUsers, createOrder).
Drop in your OpenAPI JSON/YAML or a URL. Nothing to install, no account.
A 0–100 agent-readiness score across 6 categories, with every finding listed for free.
Pay once to reveal a concrete, copy-pasteable fix for each finding.
Paste a spec. Free score, free findings. Fixes when you want them.