# APIs.io — Federated API Discovery Network > APIs.io is an agent-friendly directory of public-internet APIs. It indexes API providers, their individual APIs, schemas, event-driven specs, governance rules, and JSON-LD contexts — all served under the single `apis.io` root domain (each section at its own subpath, e.g. `/providers/`, `/apis/`, `/tags/`). APIs.io is intended for both humans and agents. Indexing, retrieval-augmented generation (RAG), grounding, and model training are explicitly permitted — see https://apis.io/terms/. Every page exposes a `` pointing at an RFC 9727 linkset that enumerates the indexed APIs, and robots.txt permits search, ai-input, and ai-train via Cloudflare Content-Signals and AIPREF Content-Usage. The catalog currently indexes roughly 27554 providers, 133606 APIs, and 127117 schemas. ## For agents: API, MCP & skills The entire catalog is queryable programmatically — start here: - [REST API](https://apis.io/api/v1): read-only discovery API. Contract: [OpenAPI](https://raw.githubusercontent.com/api-evangelist/apis-io/refs/heads/main/openapi/apis-io-v1-openapi.yml). Docs: https://apis.io/developer/ - [MCP server](https://apis.io/mcp): Streamable HTTP Model Context Protocol server, led by `apis_io_search`, with guided prompts for every job (`find_api`, `integrate_provider`, `agent_readiness_scan`, `improve_my_score`). Call `tools/list` for the current tool set. Card: https://apis.io/.well-known/mcp/server-card.json - [Agent skills](https://apis.io/skills/): 20 skills covering discovery, adoption, agent readiness, scoring, and market research. Index at https://apis.io/.well-known/agent-skills/index.json - [Claude Code plugin](https://apis.io/skills/): all the skills plus the MCP server in one install — `/plugin marketplace add api-search/network` then `/plugin install apis-io@apis-io` - [api-catalog](https://apis.io/.well-known/api-catalog): RFC 9727 linkset of every machine-readable surface - [Plans](https://apis.io/developer/plans/): free + paid tiers (industries/regions/ratings need Pro+) ## Onboarding, identity and payment — the trust layer An agent can go from cold discovery to an AUTHORIZED call here with no human at any step: read the catalog, read the protected-resource metadata, register or sign, and take a credential. It cannot yet PAY without a human — per-call settlement is built and not switched on, and saying otherwise would be the one overclaim this whole layer exists to avoid. These five pages describe that path; the machine-readable half is linked from each. - [The Trust Layer](https://apis.io/trust/): how discovery, identity, authentication, payment and accountability fit together as one system — and which parts are live, which are built but not switched on, and which are deliberately not built. Start here if you want the shape before the detail. - [Onboarding](https://apis.io/onboarding/): the four steps from a cold start to an authorized call, and the five doors in — `open` (no credential), `browser-oauth`, `dcr`, `cimd`, and `aauth` for an agent with nobody in the loop. Machine twin: https://apis.io/.well-known/api-onboarding - [KYA — Know Your Agent](https://apis.io/kya/): how a consumer tells a provider who it is, so onboarding can happen without a signup form. Six facets, 26 checks, four grades. Two numbers, never one — *disclosure* is what you told us, *verification* is what we independently confirmed. Graded facts, never a trust verdict: the provider decides, weighted by its own priorities. - [Blueprint](https://apis.io/blueprint/): every machine-readable document APIs.io serves, the specification behind each, and what each is worth on the Kin Score. Written to be copied — 50 of the 139 Agent Readiness points come from publishing documents about an API you already have. - [Rating](https://apis.io/rating/): the public rubric behind all of it. Every score is recomputable, and correcting an error we made is free. Discovery documents an agent can fetch directly, all keyless: - https://apis.io/.well-known/api-catalog — RFC 9727 + RFC 9264 linkset of every surface - https://apis.io/.well-known/oauth-protected-resource — RFC 9728, names the authorization server - https://apis.io/.well-known/oauth-authorization-server — RFC 8414, carries `registration_endpoint` - https://apis.io/.well-known/api-onboarding — API Onboarding Descriptor, including what we require of a consumer - https://apis.io/.well-known/aauth-resource.json — agent authentication by signature, no prior registration - https://apis.io/.well-known/http-message-signatures-directory — RFC 9421 keys - https://apis.io/.well-known/mcp/server-card.json — the MCP server card - https://apis.io/.well-known/agent-skills/index.json — the agent skills index ## Core entry points - [APIs.io home](https://apis.io/): Search across the whole network by provider, schema, or name - [About APIs.io](https://apis.io/about/): What the network is and how it is built - [Add an API](https://apis.io/add/): How to submit a provider profile - [Stack](https://apis.io/stack/): Assemble a watchlist of the providers you depend on and export it as an APIs.json 0.21 `apis.yml` — each provider carrying its Kin Score, agent readiness, and published resources. Browser-only (one cookie), no account. - [Blog](https://apis.io/blog/): Network updates, provider profiles, weekly roundups - [Insights](https://apis.io/insights/): Cross-network reports (top providers, top tags, spec-format breakdowns) - [Agent Skills](https://apis.io/agent-skills/): Agent skills published by API providers across the network ## Kin Score — the rating model, one page per measurement Every provider carries a Kin Score composite (0-100) and a separate agent-readiness score. The rubric is open: every check, its points, and its predicate are published, and each measurement has its own page carrying its definition, its checks, the catalog-wide distribution, and the top-scoring providers (capped at 500, with the true population stated). - [Kin Score rubric](https://apis.io/rating/): bands, facets, all checks, the regulatory applicability map, and how a score is computed - [Contract Quality](https://apis.io/rating/facets/contract-quality/): composite facet, 25% weight, 42 checks - [Access Clarity](https://apis.io/rating/facets/commercial-clarity/): composite facet, 20% weight, 9 checks - [Developer Ergonomics](https://apis.io/rating/facets/developer-ergonomics/): composite facet, 20% weight, 13 checks - [Operational Transparency](https://apis.io/rating/facets/operational-transparency/): composite facet, 13% weight, 9 checks - [Contract Governance](https://apis.io/rating/facets/governance/): composite facet, 12% weight, 6 checks - [Discoverability](https://apis.io/rating/facets/discoverability/): composite facet, 10% weight, 12 checks - [Regulatory Posture](https://apis.io/rating/facets/regulatory/): composite facet, 15% weight (conditional), 18 checks - [Open Source Surface](https://apis.io/rating/facets/open-source/): composite facet, 10% weight (conditional), 4 checks - [Create-or-Update Ergonomics](https://apis.io/rating/facets/upsert/): composite facet, 10% weight (conditional), 4 checks - [Machine-Readable Contract](https://apis.io/rating/dimensions/spec-presence/): agent-readiness dimension, 18 points - [Agentic Access Contract](https://apis.io/rating/dimensions/agentic-access/): agent-readiness dimension, 10 points - [Documented Reversibility](https://apis.io/rating/dimensions/reversibility-documented/): agent-readiness dimension, 6 points - [MCP Server](https://apis.io/rating/dimensions/mcp-server/): agent-readiness dimension, 12 points - [Machine-Readable Auth](https://apis.io/rating/dimensions/auth-clarity/): agent-readiness dimension, 10 points - [Idempotency](https://apis.io/rating/dimensions/idempotency/): agent-readiness dimension, 9 points - [Stable Error Semantics](https://apis.io/rating/dimensions/error-semantics/): agent-readiness dimension, 8 points - [Request/Response Examples](https://apis.io/rating/dimensions/openapi-examples/): agent-readiness dimension, 7 points - [Rate-Limit Signaling](https://apis.io/rating/dimensions/rate-limit-signal/): agent-readiness dimension, 7 points - [Typed Event Surface](https://apis.io/rating/dimensions/event-surface-described/): agent-readiness dimension, 6 points - [Agent Skills](https://apis.io/rating/dimensions/agent-skills/): agent-readiness dimension, 5 points - [Well-Known Catalog](https://apis.io/rating/dimensions/well-known-catalog/): agent-readiness dimension, 4 points - [Consent & Bot Identity](https://apis.io/rating/dimensions/consent-identity/): agent-readiness dimension, 3 points - [A2A Agent Card](https://apis.io/rating/dimensions/agent-card/): agent-readiness dimension, 8 points - [Dry-Run / Simulate Mode](https://apis.io/rating/dimensions/dry-run-mode/): agent-readiness dimension, 4 points - [Delegated User Identity](https://apis.io/rating/dimensions/delegated-identity/): agent-readiness dimension, 6 points - [Protected Resource Metadata](https://apis.io/rating/dimensions/protected-resource-metadata/): agent-readiness dimension, 5 points - [Registration Without a Human](https://apis.io/rating/dimensions/dynamic-client-registration/): agent-readiness dimension, 6 points - [Agentic Commerce Well-Known Document](https://apis.io/rating/dimensions/agentic-commerce/): agent-readiness dimension, 5 points ## Sister sites in the APIs.io network - [providers.apis.io](https://apis.io/providers/): API provider profiles — one page per organization (Stripe, GitHub, FDA, …) - [apis.apis.io](https://apis.io/apis/): Individual API records — one page per published API, with OpenAPI links - [industries.apis.io](https://apis.io/industries/): Industry-vertical index — every provider mapped to one or more of 29 industries (financial services, healthcare, manufacturing, transportation, etc.) - [regions.apis.io](https://apis.io/regions/): Sales-region index — providers grouped by macro sales region (North America, LATAM, EMEA, APAC, Greater China, ANZ, …) - [agent-skills.apis.io](https://apis.io/agent-skills/): Provider Agent Skill index — 800+ SKILL.md files across 80+ providers, rendered inline with per-skill and per-provider rollup pages - [mcp.apis.io](https://apis.io/servers/): Official MCP server index — 230+ Model Context Protocol servers across 220+ providers, with install hints and source links - [schemas.apis.io](https://apis.io/schemas/): JSON Schemas extracted from the indexed APIs - [asyncapi.apis.io](https://apis.io/asyncapis/): AsyncAPI event-driven specifications - [json-ld.apis.io](https://apis.io/jsonld/): JSON-LD contexts and semantic vocabularies - [rules.apis.io](https://apis.io/rules/): Spectral governance rules - [tags.apis.io](https://apis.io/tags/): Tag index across the network - [examples.apis.io](https://examples.apis.io/): API usage examples - [developer.apis.io](https://apis.io/developer/): Developer portal and feeds - [plans.apis.io](https://apis.io/plans/): API pricing plan profiles - [rate-limits.apis.io](https://apis.io/rate-limits/): API rate-limit profiles - [finops.apis.io](https://apis.io/finops/): API FinOps profiles ## Machine-readable feeds - [/.well-known/api-catalog](https://apis.io/.well-known/api-catalog): RFC 9727 linkset for the apis.io hub - [providers.apis.io api-catalog](https://apis.io/.well-known/api-catalog): Linkset of every indexed provider - [apis.apis.io api-catalog](https://apis.io/.well-known/api-catalog): Linkset of every indexed API - [APIs.json profile](https://apis.io/apis.json): Self-describing APIs.json document for the network - [sitemap.xml](https://apis.io/sitemap.xml): Per-subdomain sitemaps live at `/sitemap.xml` - [search-index.json](https://apis.io/search-index.json): Front-end search index used by the hub homepage ## Related projects - [APIs.json](https://apisjson.org): The machine-readable API discovery format that every profile in this network uses - [API Commons](https://apicommons.org): Standard vocabulary of common operational properties referenced from APIs.json - [api-evangelist](https://github.com/api-evangelist): Upstream Git-versioned provider profile repos this network is built from ## License and usage All content on APIs.io is published as a public catalog of API metadata, intended for both human and machine consumption. Search indexing, AI grounding/RAG, and AI training are explicitly permitted. See https://apis.io/terms/ for the full terms.