How do I install the Web To Markdown skill?

Run npx skillmds@latest add rungchan2/web-to-markdown in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Web To Markdown safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: executes scripts, makes network calls, reads secrets. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Web To Markdown?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Web To Markdown free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Web To Markdown?

rungchan2 (@rungchan2) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Web To Markdown by rungchan2 — Agent Skill

Name: Web To Markdown
Rating: 4.5 (4 reviews)
Author: rungchan2

Web To Markdown

Fetch web pages and save them as clean Markdown files into a docs/ directory. Use when the user provides a URL and wants its content saved as a local .md file for reference — especially documentation pages, API references, blog posts, or technical guides. Triggers on requests like "save this page as markdown", "fetch this docs page", "grab this URL into docs/", or when given a URL with intent to capture its content locally. Supports Cloudflare Markdown for Agents (Accept text/markdown) for higher quality output on supported sites, with automatic HTML-to-Markdown fallback for all other sites. Configurable via CLOUDFLARE_API_KEY environment variable for authenticated Cloudflare access.

rungchan2 Updated Sep 10, 2026

File contents

Web to Markdown

Fetch a URL and save its content as a clean .md file in the project's docs/ directory.

Quick Start

Run the bundled script:

python3 <skill-path>/scripts/fetch_markdown.py <url> [--output <path>]

The script automatically:

Tries Accept: text/markdown header (Cloudflare Markdown for Agents)
Falls back to HTML fetch + conversion if markdown not available
Saves to docs/<url-derived-name>.md by default

Options

Flag	Description	Default
`--output`, `-o`	Custom output path	`docs/<name>.md`
`--api-key`	Cloudflare API key	`$CLOUDFLARE_API_KEY` env
`--zone-id`	Cloudflare Zone ID	`$CLOUDFLARE_ZONE_ID` env

Usage Patterns

Single page

python3 <skill-path>/scripts/fetch_markdown.py "https://react.dev/reference/react/useState"
# → docs/reference_react_useState.md

Custom output path

python3 <skill-path>/scripts/fetch_markdown.py "https://docs.example.com/api" --output docs/api/example-api.md

Multiple pages (run sequentially)

python3 <skill-path>/scripts/fetch_markdown.py "https://docs.example.com/getting-started" --output docs/getting-started.md
python3 <skill-path>/scripts/fetch_markdown.py "https://docs.example.com/api-reference" --output docs/api-reference.md

With Cloudflare API key

export CLOUDFLARE_API_KEY="your-key"
python3 <skill-path>/scripts/fetch_markdown.py "https://example.com/docs"

Conversion Strategy

Cloudflare sites: If the site has Markdown for Agents enabled, the server returns clean markdown directly — no client-side parsing needed. This produces the highest quality output.
All other sites: HTML is fetched and converted locally. The script extracts <main>, <article>, or <body> content, strips nav/footer/scripts, and converts to markdown. Quality is good for documentation sites but may include some noise for complex layouts.

Output Format

Each saved file includes source metadata:

<!-- Source: https://example.com/docs/page -->
<!-- Fetched: 2026-02-17 14:30 -->

# Page Title
...content...

Notes

No external Python dependencies required (uses only stdlib)
Cloudflare API key is optional — the Accept: text/markdown header works without authentication on many Cloudflare-proxied sites that have enabled the feature
For JavaScript-heavy SPAs that render client-side, the HTML fallback may capture minimal content. Consider using a browser-based approach for those sites.

Resources

scripts/

fetch_markdown.py — Main fetch and conversion script. Zero dependencies, runs with Python 3.10+.

---name: web-to-markdowndescription: Fetch web pages and save them as clean Markdown files into a docs/ directory. Use when the user provides a URL and wants its content saved as a local .md file for reference — especially documentation pages, API references, blog posts, or technical guides. Triggers on requests like "save this page as markdown", "fetch this docs page", "grab this URL into docs/", or when given a URL with intent to capture its content locally. Supports Cloudflare Markdown for Agents (Accept text/markdown) for higher quality output on supported sites, with automatic HTML-to-Markdown fallback for all other sites. Configurable via CLOUDFLARE_API_KEY environment variable for authenticated Cloudflare access.---# Web to MarkdownFetch a URL and save its content as a clean `.md` file in the project's `docs/` directory.## Quick StartRun the bundled script:```bashpython3 <skill-path>/scripts/fetch_markdown.py <url> [--output <path>]```The script automatically:1. Tries `Accept: text/markdown` header (Cloudflare Markdown for Agents)2. Falls back to HTML fetch + conversion if markdown not available3. Saves to `docs/<url-derived-name>.md` by default## Options| Flag | Description | Default ||------|-------------|---------|| `--output`, `-o` | Custom output path | `docs/<name>.md` || `--api-key` | Cloudflare API key | `$CLOUDFLARE_API_KEY` env || `--zone-id` | Cloudflare Zone ID | `$CLOUDFLARE_ZONE_ID` env |## Usage Patterns### Single page```bashpython3 <skill-path>/scripts/fetch_markdown.py "https://react.dev/reference/react/useState"# → docs/reference_react_useState.md```### Custom output path```bashpython3 <skill-path>/scripts/fetch_markdown.py "https://docs.example.com/api" --output docs/api/example-api.md```### Multiple pages (run sequentially)```bashpython3 <skill-path>/scripts/fetch_markdown.py "https://docs.example.com/getting-started" --output docs/getting-started.mdpython3 <skill-path>/scripts/fetch_markdown.py "https://docs.example.com/api-reference" --output docs/api-reference.md```### With Cloudflare API key```bashexport CLOUDFLARE_API_KEY="your-key"python3 <skill-path>/scripts/fetch_markdown.py "https://example.com/docs"```## Conversion Strategy1. **Cloudflare sites**: If the site has Markdown for Agents enabled, the server returns clean markdown directly — no client-side parsing needed. This produces the highest quality output.2. **All other sites**: HTML is fetched and converted locally. The script extracts `<main>`, `<article>`, or `<body>` content, strips nav/footer/scripts, and converts to markdown. Quality is good for documentation sites but may include some noise for complex layouts.## Output FormatEach saved file includes source metadata:```markdown# Page Title...content...```## Notes- No external Python dependencies required (uses only stdlib)- Cloudflare API key is optional — the `Accept: text/markdown` header works without authentication on many Cloudflare-proxied sites that have enabled the feature- For JavaScript-heavy SPAs that render client-side, the HTML fallback may capture minimal content. Consider using a browser-based approach for those sites.## Resources### scripts/- `fetch_markdown.py` — Main fetch and conversion script. Zero dependencies, runs with Python 3.10+.

Frequently asked questions