1Fetch raw content from a URL. Fast and lightweight - no AI processing.
 2
 3<when_to_use>
 4Use Fetch when:
 5- Need raw HTML, JSON, or text from a URL
 6- Accessing API endpoints directly
 7- Want content without interpretation
 8- Saving tokens (no AI processing)
 9
10Do NOT use Fetch when:
11- Need to extract specific information → use `agentic_fetch`
12- Need to analyze or summarize content → use `agentic_fetch`
13- Want to search the web → use `agentic_fetch` without URL
14- Downloading binary files → use `download`
15</when_to_use>
16
17<parameters>
18- url: URL to fetch (required)
19- format: "text", "markdown", or "html" (required)
20- timeout: Seconds to wait (optional, max 120)
21</parameters>
22
23<format_guide>
24- `text`: Plain text, best for APIs or simple content
25- `markdown`: Converted from HTML, good for documentation
26- `html`: Raw HTML structure
27</format_guide>
28
29<limits>
30- Max response: 5MB
31- HTTP/HTTPS only
32- No authentication or cookies
33- Some sites block automated requests
34</limits>
35
36<example>
37Fetch API response:
38```
39url: "https://api.github.com/repos/owner/repo"
40format: "text"
41```
42
43Fetch documentation as markdown:
44```
45url: "https://docs.example.com/api"
46format: "markdown"
47```
48</example>