web_fetch
Fetch and read full content from any URL, including bot-protected sites and package registries. Supports pagination with cursors for large documents.
Instructions
Fetch and read content from a URL via EnriProxy's multi-tier fetch service.
When to use:
When you need to read the full content of a webpage.
When you need to access documentation, articles, or code files.
When simpler fetch methods fail due to anti-bot protection.
Features:
Package registry API detection (npm, PyPI)
Raw file fetch (GitHub raw, HuggingFace)
Robust fetching for static, dynamic, and bot-protected sites (best-effort)
Automatic fallback across multiple retrieval strategies (details intentionally not exposed)
Notes:
Provide the full URL including protocol (https://).
Content is limited by the
max_charsparameter (default: 200000).If the result is truncated and includes a
cursor, callweb_fetchagain withcursor+offset+limitto read more without re-fetching.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| url | No | Full URL to fetch (http:// or https://). | |
| cursor | No | Opaque cursor returned by a previous `web_fetch` call for pagination. | |
| prompt | No | Optional hint describing what you want to extract (the tool returns fetched content; it does not generate an AI summary). | |
| max_chars | No | Maximum content length (default: 200000). | |
| offset | No | Cursor read offset in characters (default: 0). | |
| limit | No | Cursor read limit in characters (default: max_chars). |