| name | jina-reader |
| description | Fetch clean, readable content from any URL using Jina Reader. Use when the user wants to read, summarize, or analyze a web page, article, or blog post. |
| allowed-tools | Bash(curl *) |
| argument-hint | <url> |
Jina Reader
Fetch clean, readable content from any URL using Jina Reader.
Usage
curl -sL "https://r.jina.ai/<url>"
For example:
curl -sL "https://r.jina.ai/https://example.com/article"
Notes
- Works with most web pages, articles, and blog posts
- Returns clean markdown-formatted content stripped of navigation, ads, and other clutter
- Do NOT use for GitHub URLs — use
gh CLI or raw.githubusercontent.com instead
- If Jina Reader fails for a specific URL, fall back to WebFetch
What to do with the output
After fetching the content, ask the user what they'd like to do with it. Common tasks:
- Summarize the article
- Extract key points or action items
- Answer questions about the content
- Compare with other articles