--- title: "anymd — The web context layer for AI agents" url: "https://md.harnesskit.cc/" --- # anymd > The web context layer for AI agents. ## Turn the web into context your agents can use. anymd is the web context layer for AI agents: it reads public web content, normalizes it into structured Markdown, remembers what matters in a private searchable library, and makes it available anywhere through MCP, API and CLI. Try it now: prefix any public URL with `anymd.cc/`, no account needed (up to 50 reads a day). ``` https://md.harnesskit.cc/https://stephango.com/saw ``` Connect your agent over MCP: `https://md.harnesskit.cc/mcp` — [setup](https://md.harnesskit.cc/docs/mcp.md). Open source · Self-hostable · Respects robots.txt · No paywall or login bypass. ## The problem Raw HTML is noisy agent context. A typical page is mostly menus, trackers, pop-ups, related posts and inline styles, so an agent that reads it burns tokens, context and accuracy on noise. ## Read A source-aware reader extracts the content and normalizes it into structured Markdown with YAML metadata. Headings, links, tables, code and footnotes survive; clutter does not. Dedicated readers handle sources such as GitHub, YouTube, Reddit, Hacker News, X and documents. ## Remember Your agents shouldn't read the same web twice. Every source they read while signed in becomes reusable private context in a library searchable by keyword, phrase, meaning, or through MCP: - **BM25** relevance ranking over titles, text, domains and tags - **Full-text** search with exact phrases, boolean operators and prefixes - **Semantic** search with multilingual bge-m3 embeddings - **Hybrid** search fusing BM25 and semantic results with Reciprocal Rank Fusion - **Query fan-out**, rewriting a query into variants - Optional **Jev** tie-breaking when the top results are too close to call Library search and recall never cost credits. ## Use anywhere - **MCP:** `https://md.harnesskit.cc/mcp` (Streamable HTTP, OAuth or API key). Tools include `read_url`, `search_library` and `get_document` — [docs](https://md.harnesskit.cc/docs/mcp.md) - **REST API:** `POST https://md.harnesskit.cc/api/v1/convert` with `Authorization: Bearer amd_…` — [docs](https://md.harnesskit.cc/docs/api.md) - **CLI:** `npm i -g https://cdn.anymd.cc/cli/anymd-cli-latest.tgz` — [docs](https://md.harnesskit.cc/docs/cli.md) - **WebMCP:** tools exposed to in-browser agents on anymd.cc — [docs](https://md.harnesskit.cc/docs/webmcp.md) - **URL API:** `GET https://md.harnesskit.cc/` returns Markdown, `?format=json` for JSON — [docs](https://md.harnesskit.cc/docs/url.md) ## Responsible by design - **Requested URLs only.** anymd reads the URLs a user or their agent asks for. It does not crawl or discover pages on its own. - **Public content only.** Only pages that anyone can reach on the open web, without an account. - **No bypassing access.** No login, paywall or CAPTCHA bypass. Terms forbid users from trying, too. - **Respects robots.txt.** Checked before anymd fetches a page itself, for the anymd token or *. Disallowed pages are refused. - **Domain opt-out.** Site owners can block their domain on every channel, including cached results. - **Per-site rate limits.** Each site has a shared per-minute fetch budget, so agents never hammer a server. - **Private library.** What your agents read is saved to your account only. Never shared, never used to train models. - **Takedown process.** Abuse reports and takedown requests go to a human at hello@digitop.ai. Site owners can opt out or request a takedown: https://md.harnesskit.cc/legal/abuse.md ## Supported sources - **Any web page** — Articles, docs, blogs — clutter removed - **X / Twitter** — Posts, quotes, media, engagement stats - **Facebook** — Public posts, captions and media; comments on request - **Instagram** — Public posts, reels and media; comments on request - **Threads** — Public posts and carousel media; comments on request - **LinkedIn** — Public posts, media and comments on request - **YouTube** — Title, channel and a timestamped transcript - **GitHub** — READMEs, issues, PRs and discussions - **Reddit** — Threads with nested comments - **Hacker News** — Stories plus the top of the discussion - **PDF** — Text layout kept, tables preserved - **DOCX · XLSX · CSV** — Office files and sheets as Markdown tables - **Images** — Vision model describes and transcribes Planned: Audio & podcasts, Any video (Whisper), TikTok, Notion, Google Docs, EPUB. ## Pricing Pay when your agents learn something new. Reusing what they already know is free: processing a new source uses credits by complexity, while cached reads and library search cost nothing. | Plan | Price | Credits / month | Beyond included | |---|---|---|---| | Free | $0 | 500 | hard cap | | Pro | $9/mo ($7/mo yearly) | 10,000 | $1 per extra 1k | | Scale | $49/mo ($39/mo yearly) | 100,000 | $0.6 per extra 1k | | Enterprise | Custom | 1,000,000 | $0.4 per extra 1k | | Source | Credits | |---|---| | Web page, GitHub, Reddit, Hacker News, X post | 1 | | YouTube video with transcript | 3 | | PDF, DOCX, XLSX, CSV (per file) | 3 | | Image → Markdown (vision description) | 5 | | Facebook, Instagram, Threads or LinkedIn post | 10 | | Each additional X thread post | 1 | | Each started batch of 20 retrieved comments/replies | 10 | | Each successfully analyzed article image | 5 | | YouTube video stored on the anymd CDN (opt-in, charged when ready) | 20 | | AI analysis of a stored YouTube video, per started minute (plus 10 per video; opt-in, charged when ready) | 20 | | Social search: each page of X, Facebook, Instagram or Threads results | 10 | | Social search: each page of LinkedIn results | 100 | | Library search, reads, MCP reads | 0 | Launch offer: 30% off Pro and Scale with code `LAUNCH30` until 2026-10-31. Full details: https://md.harnesskit.cc/pricing.md ## Self-hosting anymd is open source (MIT) and deploys as a single Cloudflare Worker: https://md.harnesskit.cc/docs/self-host.md · https://github.com/digitopvn/anymd ## Built by Duy Nguyen (/zuey/), Founder, Digitop.ai. ## FAQ ### Is anymd free? Yes. Anyone can prefix a URL with anymd.cc/ — no account, up to 50 reads a day. A free account gives you 500 credits a month plus a private, searchable context library. Paid plans start at $9/month. ### What is a credit? Credits pay for new source processing, priced by complexity: a web page is 1 credit, a YouTube transcript or a PDF/Office file is 3, an image is 5. Reusing what your agents already know is free: cached reads, library search and MCP recall never cost credits. ### How is this different from fetching the HTML? anymd is a source-aware reader: a proven content-extraction engine plus dedicated paths for sources such as GitHub, YouTube, Reddit, Hacker News, X and documents. Your agent gets the content, not the nav, cookie banners, ads or related posts, as structured Markdown with metadata, headings, links, tables, code and footnotes intact. ### How do my agents use it? Connect the MCP server at anymd.cc/mcp (OAuth or API key) and your agent gets read_url, search_library and get_document tools. Apps use the REST API, pipelines use the CLI, and in-browser agents get WebMCP tools on anymd.cc. ### What happens to what my agents read? Signed-in reads are saved to your private library so you and your agents can search them later. Nothing is shared, nothing is used to train models, and you can delete any document — or your whole account — at any time. ### Does anymd crawl websites or get past paywalls? No. anymd only reads URLs that a user or their agent asks for, and only content that is publicly reachable. It follows robots.txt, honours domain opt-outs, limits how often it hits each site, and never bypasses logins, paywalls or CAPTCHAs. ### Is it open source? Yes, MIT-licensed at github.com/digitopvn/anymd. You can self-host it on your own Cloudflare account. ### What if I run out of credits? On Free you wait for next month or upgrade. On Pro and Scale you are never blocked: extra usage is billed at $1 (Pro) or $0.60 (Scale) per 1,000 credits. ### Can I get a refund? Yes — 14-day money-back on your first subscription payment if you used less than 20% of the month’s credits. See the refund policy for details.