🧰 OpenAPI Spec Scraper — API Directory → Agent-Tool / RAG Dataset

Turn the world's machine-readable APIs into a clean, LLM-agent-ready dataset. Bulk-scrape the apis.guru directory (2,500+ public API specs) — or any OpenAPI/Swagger URL — and get normalized rows, one per endpoint: method, path, summary, parameters, auth scheme, server, and tags. Perfect for building agent tool catalogs, MCP servers, and RAG knowledge bases. Monitor the directory for newly-added APIs.

No login, no anti-bot — the directory and specs are public JSON.

---

What does it do?

Most spec converters take one pasted spec. This actor works on the bulk directory axis: point it at apis.guru (search or take the top N), and it fetches each API's OpenAPI/Swagger spec and normalizes it into ready-to-load rows.

• 📚 2,500+ APIs in the apis.guru directory (or your own spec URLs).
• 🔧 One row per endpoint (agent-tool ready): method, path, operation_id, summary, parameters, auth, tags, server — or one row per API with an endpoints[] array.
• 🔔 Monitor mode — alert when new APIs are added to the directory.

---

What data does it extract?

Per endpoint (emitMode: per-endpoint): api, api_title, server, auth_schemes, method, path, operation_id, summary, tags, parameters[] ({name, in, required, type}), request_body_required, response_codes, scraped_at.

Per API (emitMode: per-api): api, title, description, version, spec_version, server, auth_schemes, endpoint_count, endpoints[].

---

Who is it for?

• 🤖 AI agent / tool builders & MCP authors assembling machine-usable tool catalogs.
• 📚 RAG / knowledge-base builders indexing API documentation as structured text.
• 🧑‍💻 Developer-tooling teams building API search, SDKs, and integrations.

---

How to use it (step by step)

1. Click Try for free.
2. Search the directory (e.g. stripe, google, weather) — or paste spec URLs / API names.
3. Pick the output shape (per-endpoint is agent-tool ready).
4. Click Start, then open the Dataset tab to view/export.
5. (Optional) set monitorMode + a Schedule to catch newly-added APIs.

Quick start

{ "searchTerm": "stripe", "emitMode": "per-endpoint" }

Bulk-directory example

{ "maxApis": 200, "emitMode": "per-endpoint", "includeParameters": true }

---

Input

Field	What it does
searchTerm	filter the apis.guru directory by name/title
apiNames	specific directory keys (e.g. stripe.com)
specUrls	any raw OpenAPI/Swagger spec URLs
emitMode	per-endpoint (agent-tool ready) · per-api
includeParameters, includeSchemas	enrich each endpoint
maxApis	cap on directory APIs (0 = all matching)
monitorMode, alertOnNewApi	recurring new-API watcher + alerts
webhookUrl, slackWebhookUrl, emailRecipients	alert channels
proxyConfiguration, requestConcurrency	proxy + parallelism

---

Output

One dataset record per endpoint (or per API). Export to JSON, CSV, Excel, HTML, or RSS — or fetch JSONL via the Apify API to load straight into an agent/RAG pipeline.

---

How much does it cost?

Pay-per-event — you pay for what you pull:

Event	What it covers	Suggested price
lot-scraped	each endpoint (or API) row returned	~$0.001 / row
monitor-run-completed	each scheduled watch run	~$0.05 / run
new-lot-detected	each newly-added API	~$0.02 / API
alert-delivered	each Slack/email/webhook push	~$0.005 / alert

(Final per-event prices are set on the actor's pricing page.)

---

Is it legal?

The apis.guru directory and the linked specs are public, openly-licensed API descriptions — no personal data. You're responsible for your use: review apis.guru's terms and each spec's own license/terms before redistributing.

---

FAQ

OpenAPI 3 and Swagger 2? Both — the normalizer handles openapi 3.x and swagger 2.0 (server, auth, parameters).

Can I use a spec that isn't in the directory? Yes — pass its URL in specUrls.

What's the best output for an AI agent? per-endpoint — each row is a self-contained tool descriptor.

How do I export? JSON / JSONL, CSV, Excel, HTML, or RSS from the Dataset tab, or via the Apify API.

---

Feedback

Want full request/response schema flattening or OpenAPI→MCP tool-manifest output? Open an issue on the actor.