Mdrip

mdrip

Use this skill when an agent needs local, reusable markdown context from web pages for implementation, debugging, or documentation work.

When to use

• You need to ingest docs/blog pages into the repository as markdown.
• You want to prefer Accept: text/markdown for Cloudflare Markdown for Agents.
• You still need good results when a site only returns HTML.
• You need raw markdown streamed to stdout for agent runtimes (for example OpenClaw) without writing local snapshot files.
• You need to call mdrip as a library from Node.js or Cloudflare Workers.

Core workflow

1. Fetch target URLs with markdown content negotiation.
2. Detect whether markdown was served (content-type: text/markdown).
3. If markdown is not served and fallback is allowed, convert HTML to markdown.
4. Save snapshots under mdrip/pages/.../index.md.
5. Update source tracking in mdrip/sources.json.
6. Return a concise summary of paths, content mode, and failures.

Commands

# fetch one page
npx mdrip <url>

# fetch many pages
npx mdrip <url1> <url2> <url3>

# strict Cloudflare markdown only (no html fallback)
npx mdrip <url> --no-html-fallback

# raw markdown to stdout only (no settings/snapshot writes)
npx mdrip <url> --raw

# inspect tracked pages
npx mdrip list --json

Programmatic usage

// Workers/agent runtimes (no filesystem writes)
import { fetchMarkdown, fetchRawMarkdown } from "mdrip";

// Node.js filesystem helpers
import { fetchToStore, fetchManyToStore, listStoredPages } from "mdrip/node";

Guardrails

• Prefer official sources and canonical URLs.
• Do not overwrite unrelated files.
• Report whether each result came from Cloudflare markdown or HTML fallback.
• If a fetch fails, include URL, HTTP status/error, and next-step retry guidance.

References

• references/workflow.md
• references/fallback-and-quality.md