# AgentFetch

> AI agent web access, done as a tool call: how to design a fetch tool, handle its errors, cap its tokens and keep fetched pages from steering your agent.

Field notes on giving AI agents web access they can actually use. A Teneo Labs project · Built on Live Web Data API.

## Guides

- [How to design a web fetch tool for AI agents](https://agentfetch.live/guides/design-a-web-fetch-tool-for-agents): Name, description, parameters, output and metadata of a web fetch tool, and how each choice changes what the model does with it inside the agent loop.
- [Agent tool error handling: a field guide for web fetch](https://agentfetch.live/guides/agent-tool-error-handling): Which web fetch failures to retry in code, which to hand to the model, and how to word the error so the agent picks a sensible next step instead of looping.
- [Indirect prompt injection from web pages: an agent's threat model](https://agentfetch.live/guides/prompt-injection-from-web-pages): Why every page an agent fetches is untrusted input, how hidden instructions and model-built URLs cause harm, and the controls that hold up in practice.
- [Token budgets for agent web pages: caps, limits and costs](https://agentfetch.live/guides/token-budgets-for-agent-web-pages): How much context a fetched page takes, why it costs again on every turn, and how to cap pages, budget fetches per task and summarize before the main agent.
- [Browser agent vs fetch: which tool each agent step needs](https://agentfetch.live/guides/browser-agent-vs-fetch): When an agent step needs a fetch tool, a search API, a browser agent or computer use, compared on what each can read, cost, failure modes and risk.

## Posts


## Built on Live Web Data API

> Live web pages as clean Markdown, text or HTML for AI agents. Send up to 10 URLs per request to POST /v1/contents; failed pages cost nothing.

- API base URL: https://livewebdata.com/api
- Auth: `Authorization: Bearer <api_key>` (keys start with 1337_live_)
- OpenAPI: https://agentfetch.live/openapi.json
- MCP server (Streamable HTTP, same bearer key): https://agentfetch.live/mcp

### Docs

- [Overview](https://agentfetch.live/docs.md): Live Web Data API turns any public URL into clean Markdown, text or HTML — fetched live through our fetch network.
- [API reference](https://agentfetch.live/docs/api-reference.md): Two endpoints. HTTP 200 is returned even when individual pages failed — check statuses[] for the per-URL outcome.
- [Examples](https://agentfetch.live/docs/examples.md): Copy-paste examples in cURL, Python and JavaScript. Replace YOUR_API_KEY with a key from the dashboard.
- [Automate requests](https://agentfetch.live/docs/automation.md): Move from clicking Run in the playground to fetching URL lists from your own code — with rate limits, retries and scheduling handled.
- [SDKs](https://agentfetch.live/docs/sdks.md): Thin, typed clients for JavaScript/TypeScript and Python — built-in retries, typed errors, one-line page fetches.
- [MCP for agents](https://agentfetch.live/docs/mcp.md): Use the Live Web Data API as a tool in Claude, Cursor, VS Code and any other MCP client — with one URL and your existing API key.
- [No-code recipes](https://agentfetch.live/docs/no-code.md): Call the API from n8n, Make or Zapier with each tool's generic HTTP step. Native integrations are on the roadmap.

### Optional

- [Full docs in one file](https://agentfetch.live/llms-full.txt)
- [Postman collection](https://agentfetch.live/live-web-data-api.postman_collection.json)
- [Example scripts (.zip)](https://agentfetch.live/live-web-data-api-examples.zip)