APIs.io serves an AI Catalog
manifest at /.well-known/ard.json on apis.io, declaring
24 agentic resources
that a discovery service can index without asking anyone’s permission.
/.well-known/ard.json — the path ARD v0.91 says a consumer MUST fetch.
specVersion in the manifest. 1.0 is current for the AI Catalog data model; anything lower is an earlier draft. ARD reads this field as transport-defined and ignores it.host-identifier-not-did-webRead-only REST discovery over the APIs.io catalog of API providers, their APIs, and the artifacts each publishes. The same contract generates the MCP server, so the two never drift. Catalog reads are free and keyless.
Learn how to read the apis.io API catalog — the v1 REST API, the MCP server, and the markdown mirror of every page — so other skills can find APIs without scraping HTML. Use this first when the user mentions apis.io or asks to discover APIs across a catalog.
Turn a domain, URL, company name, or GitHub org into an apis.io provider slug, with its quality band and composite score. Use first whenever the user names a company rather than a catalog slug — which is almost always.
Search the apis.io catalog for APIs, providers, and tags by keyword, capability, tag, or provider. Use when the user asks what APIs exist for something, which providers offer a capability, or wants to filter the catalog.
Fetch and parse the OpenAPI, AsyncAPI, GraphQL, or Postman description for an API in the apis.io catalog, given its aid or its apis.io URL. Use to inspect operations, summarize endpoints, or generate code from a real spec.
Find the right API for a task — search the apis.io catalog, inspect the strongest candidates, and recommend one with its spec and links. Use when the user asks what API to use for something, or whether an API exists for a job.
Onboard an API provider end-to-end — pull its signup, authentication, base URLs and OpenAPI from the apis.io catalog and produce concrete, runnable first-integration steps. Use when a provider is chosen and the user wants to actually wire it up.
Find providers in the apis.io catalog that publish an MCP server, with the endpoint and install method, so you can plug real tools into an agent. Use when the user wants agent-ready tools rather than raw REST.
Search the apis.io catalog for Agent Skills published by API providers — packaged SKILL.md files you can adopt directly. Use when the user wants ready-made skills for a vendor or capability rather than writing integration code.
Search the apis.io catalog for any artifact type beyond OpenAPI — Arazzo workflows, Spectral rules, pricing plans, FinOps, rate limits, GraphQL, Postman collections, JSON Schemas, examples, and more. Use when the user wants governance, commercial, or workflow artifacts rather than an API contract.
Check whether an API provider is ready for autonomous AI agents — machine-readable contracts, MCP servers, agent skills, llms.txt, agent-native access controls, and the operational signals an agent needs. Use when asked if an API is agent-ready, or to compare providers on agent readiness.
Read a provider's apis.io quality score and the Kin Score rubric, then turn it into a ranked, do-this-next punch list of what to publish to move up a band. Use when someone wants to know how their own API rates and how to improve it.
Check where an apis.io provider record's evidence came from — which artifacts the provider published first-party versus which apis.io derived — before citing a score or an artifact as fact. Use when accuracy matters, or when a provider disputes their rating.
Turn one capability into a scored, compared vendor shortlist from the apis.io catalog, with a clear recommendation. Use for a buy decision on a single capability. Pro tier.
Design a capability stack from the apis.io catalog — the best-rated provider per capability — and export it as an adoptable APIs.json. Use when the user needs several APIs at once, not one. Pro tier.
Inventory, score, and gap a set of API providers from the apis.io catalog — what they publish, how they rate, and what they are missing. Use to assess a vendor portfolio. Pro tier.
Benchmark one API provider against its closest peers facet by facet — score, artifact coverage, and where it leads or trails. Use for competitive positioning or to justify an API investment. Pro tier, with a free fallback.
Explore any slice of the API market as a scored cohort — an industry, topic area, region, or tag — with its members, their Kin Scores and agent readiness. Use to size a market, find the leaders in a category, or research a sector.
Read the demand side of the API market — what technologies a company is investing in, where it is thin, and which catalog providers are already in its stack. Use for account research, sales targeting, or to find where demand outruns supply.
Set up ongoing monitoring of the apis.io catalog using saved searches and lists — track a vendor set, get only what is new since last check, and watch for rating movement. Use for recurring vendor monitoring rather than one-off queries. Pro tier, with a free fallback.
Monitor the apis.io catalog for providers added or updated and for rating movement, across a chosen set or the whole network. Use for ongoing vendor monitoring. Pro tier.
Plain-language orientation for an agent arriving at apis.io: what the catalog holds, which surfaces read it, and what is free versus paid.
Content-Type: application/json —
no — serves application/ai-catalog+jsonAccess-Control-Allow-Origin: * —
yeshttps://apis.io/.well-known/ard.json on 2026-09-13, HTTP
200. The verbatim body is stored alongside its manifest in the
APIs.io repository
as ai-catalog/apis-io-ai-catalog.json. Nothing on this page is derived
or generated: an AI Catalog manifest is served from the publisher’s own domain or it does not
exist, which makes it — like an
A2A Agent Card — one of the few agent artifacts that cannot be
produced on a provider’s behalf. Grades are recomputed on every build; the specification is a
v0.9 draft and
these verdicts will move when it does.
Every descriptor here is available over the APIs.io API and to AI agents over MCP. Agent Discovery is not yet its own endpoint on the v1 API. Reach it through catalog search and the tag graph, or the MCP server.