Turn any URL or HTML into clean, agent-ready Markdown — local, fast, deterministic. CLI + MCP server
View on GitHub ↗Homepage ↗Turn any webpage into clean Markdown an AI agent can actually use.
distill is a local CLI (and MCP server) that takes a URL or HTML file and returns Markdown with:
- nav / ads / footers stripped - tables, code blocks, and lists kept intact - relative links resolved to absolute URLs - title / author / date in a YAML frontmatter block
No cloud API. No per-page cost. Same input → same Markdown every time.
------ Input A URL, an HTML file, or HTML on stdin Output Clean Markdown (+ optional YAML metadata) Where it runs On your machine — nothing is sent to a cloud service Who it's for Agents and developers who need docs, tables, and product pages — not just news articles
Most “reader mode” tools were built for blog posts. Agents usually need API docs, pricing tables, forum threads, and reference pages. Distill is built for that: structural fidelity first, then speed and privacy.
- CLI — distill in a shell or script - MCP server — tools distillurl, distillurls, distillhtml for local agents (Claude Code, etc.)
Published as distill-md on crates.io / npm (the name distill is taken). The installed commands are still distill and distill-mcp.
Prebuilt binaries for macOS (arm64/x64), Linux (arm64/x64), and Windows (x64) ship with every GitHub Release. See RELEASING.md for how releases are cut.
Agent-ready output (--agent-ready / --ars / -A)
Default output stays plain Markdown. With --agent-ready, Distill emits JSON with three layers agents can use without re-parsing a flat blob:
From the project README.
Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:
[](https://mcp.liqiwa.com/s/gokulnair2001--distill.html)
Ask for the symbol, not the file. Token-efficient MCP server giving AI coding agents symbol-level reads of TS/JS, Rust, Python, Java, Go, JSON, and Markdown
menot-you/n-memoryMemory your agent can trust — every answer carries its source; when it doesn't know, it says so. Hermetic, local, MCP.
Role1776/mcp-retrievalFast, idiomatic Go MCP server for web search, image search, and page scraping to Markdown. Runs locally, no API keys, free.
Smitner-Studio/facetRender agent-written Markdown as a live local page, and send the human's click back to the agent as typed data. Human-in-the-loop approval gates for coding agents.
montezuma-p/harkenYour agent cannot listen to audio. harken fixes that — locally. MCP server + CLI, one 13 MB binary, fully offline: no Python, no ffmpeg, no API key.
ieranama/discomcpTeach any AI agent how you use an MCP server – Soft-Landing to your AI integrations
The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.