fetch-url
Fetch a single URL and print its content as Markdown. Use when you need to read a web page, documentation page, or API reference without indexing it. Accepts http(s):// and file:// URLs and writes the result to stdout.
How do I install this agent skill?
npx skills add https://github.com/arabold/docs-mcp-server --skill fetch-urlIs this agent skill safe to install?
- Gen Agent Trust Hubpass
The skill allows the agent to fetch content from external URLs and convert it to Markdown using a vendor-provided package. The primary security consideration is indirect prompt injection, as the agent will ingest untrusted content from the web which could contain hidden instructions designed to manipulate the agent's logic.
- Socketpass
No alerts
- Snykfail
Risk: HIGH · 3 issues
What does this agent skill do?
Fetch URL
Fetch one URL and print its content as Markdown. Nothing is added to the documentation index.
When to use
- You need to read a single web page as Markdown.
- You want to look at a documentation page before deciding whether to index it.
- You need the text of a URL for summarizing or analysis.
To index content for repeated searching, use the docs-manage skill
(scrape command) instead.
Command
npx @arabold/docs-mcp-server@latest fetch-url <url> [options]
<url> is an http://, https:// or file:// URL.
| Flag | Default | Description |
|---|---|---|
--no-follow-redirects | follows redirects | Fail on an HTTP redirect instead of following it |
--scrape-mode auto|fetch|playwright | auto | How HTML pages are processed |
--header "Name: Value" | Custom HTTP header (repeatable) | |
--quiet | Log errors only | |
--verbose | Log debug output |
Scrape modes
| Mode | Behavior |
|---|---|
auto | Default. Renders HTML pages in headless Chromium, like playwright. |
playwright | Renders HTML pages in headless Chromium, so content built by JavaScript is included. |
fetch | Converts the HTML as the server sent it. Faster, but misses content built by JavaScript. |
In every mode, a site that refuses the plain request (HTTP 403 or 429) is fetched again through the browser.
Examples
# Fetch a documentation page
npx @arabold/docs-mcp-server@latest fetch-url https://react.dev/reference/react/useEffect
# Send an auth header
npx @arabold/docs-mcp-server@latest fetch-url https://docs.example.com/api \
--header "Authorization: Bearer tok_xxx"
# Skip browser rendering for a static page
npx @arabold/docs-mcp-server@latest fetch-url https://example.com/docs/page.html --scrape-mode fetch
# Fail instead of following a redirect
npx @arabold/docs-mcp-server@latest fetch-url https://example.com/old-page --no-follow-redirects
Output
The content goes to stdout. HTML pages and documents such as PDF or Word
files are converted to Markdown. JSON, XML and plain text are printed
unchanged. The --output flag has no effect on this command.
Logs go to stderr. Outside a terminal, only errors are logged; set
LOG_LEVEL=INFO or pass --verbose to see progress messages.
On failure the command exits with a non-zero status and prints
Unable to fetch or process the URL ... on stderr. Run it again with
--verbose to see the cause. By default the server refuses private network
addresses (such as localhost or 10.x.x.x) and local files outside the
user's Documents folder; the user can allow more in the server configuration
(scraper.security).
Tips
- Save the output to a file:
npx @arabold/docs-mcp-server@latest fetch-url <url> > page.md - After
search(docs-search skill) returns a relevant URL, fetch the page to read all of it. - For pages behind authentication, pass cookies or tokens with
--header.
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/arabold/docs-mcp-server/fetch-url">View fetch-url on skillZs</a>