| name | url-fetch |
| description | Fetch HTTP or HTTPS URLs, preview response metadata and text, or save the body to disk. Use for lightweight web/API retrieval when JavaScript execution is not required.
|
| version | 1.0.0 |
| author | Zijian Guo |
| license | MIT |
| platforms | ["linux","macos"] |
| prerequisites | {"commands":["python3"]} |
| metadata | {"hermes":{"tags":["agent-tools","http","web","api","fetch"]}} |
url-fetch - HTTP fetch utility
Fetch a URL with Python standard-library networking.
When to Use
- User asks to retrieve a raw URL or API endpoint.
- You need to save a remote file.
- You need a quick text preview and status code.
Script
~/.hermes/skills/url-fetch/scripts/url_fetch.py
Options
| Option | Default | Description |
|---|
url | required | HTTP or HTTPS URL |
-o, --output | none | Save response body to a file |
--max-chars | 4000 | Maximum characters to print when not saving |
--timeout | 20 | Request timeout in seconds |
--header | repeatable | Extra header as Name: value |
Pitfalls
- This does not execute JavaScript. Use browser automation for JS-heavy pages.
- Secrets passed in headers may be visible in shell history.