llmtomd

LLMtoMD is the memory layer for AI coding agents.

CommunityOnlineNo sign-inGlobalFreeCan modify data

What it can do

  • Convert Url: Convert a web page or a publicly reachable file URL into clean, AI-ready Markdown. Call this when the user gives you a link (article, docs page, or a PDF/Office/image file URL) and wants
  • List Documents: List the documents in the user's LLMtoMD library, most recent first. Call this to discover what the user has stored, or to find a document_id before reading, searching, or asking about
  • Get Document: Fetch a stored document's full Markdown by its document_id. Use the document_id returned by convert_url or list_documents. Call this when the user wants to read, summarize, or work with

What data it sees

Do you need an account

No: the server works without sign-in

LLMtoMD is the memory layer for AI coding agents. It converts any document — PDF, DOCX, slides, spreadsheets, images, audio, even whole websites — into clean, structured Markdown, then exposes it over MCP so your agent can search your FRDs, specs, and API docs on demand instead of re-reading (or forgetting) them.

On a real 50-page spec, answering a question via retrieval used ~97% fewer tokens than holding the whole doc in context — and returned the exact, cited requirement.

Tools

This server is OAuth-secured (it serves each user's private library), so a directory's live "Tools" tab can't enumerate it without signing in — the tools are listed here instead:

  • list_documents / get_document — browse and read your library
  • search_documents — semantic search across your docs
  • ask_documents — RAG answer with citations
  • convert_url — turn any web page into Markdown
  • save_note — write a decision/spec back into the knowledge base
  • list_collections — focus on one project

Connect (OAuth — no token to paste)

  • Cursor / VS Code: add the server URL https://mcp.llmtomd.com/mcp
  • Claude Code: claude mcp add --transport http llmtomd https://mcp.llmtomd.com/mcp
  • Also Claude, Antigravity, and API-key for OpenAI / LangChain / LlamaIndex. Full guides: https://llmtomd.com/integrations

Free tier included (MCP on every plan) → https://llmtomd.com

How to connect

How to connect llmtomd to Claude

The server address to paste into Claude settings:

https://llmtomd--support-0vd1.run.tools

  1. Open Claude (claude.ai or the desktop app).
  2. Go to Settings → Connectors.
  3. Click Add custom connector.
  4. Paste the server address copied below into Remote MCP server URL and click Add.
  5. In a chat, click + → Connectors and switch the new connector on.

Custom connectors are available on Free, Pro, Max, Team and Enterprise plans (Free is limited to one). On Team and Enterprise an organization Owner adds the connector first under Organization settings → Connectors.

Authorization

After you click Add, a sign-in window for the service opens. Sign in with your own account and approve access. Claude never sees your password.

How to connect llmtomd to ChatGPT

The server address to paste into ChatGPT settings:

https://llmtomd--support-0vd1.run.tools

  1. Open ChatGPT in a browser (chatgpt.com). A Plus, Pro, Business, Enterprise or Edu plan is required.
  2. Turn on developer mode once: Settings → Apps → Advanced settings → Developer mode.
  3. Open Settings → Connectors and click Create.
  4. Fill in the form: Name (anything), Description (one line about what the service does), MCP server URL (copy it below).
  5. Under Authentication choose OAuth if the service requires sign-in, otherwise None. Click Create.
  6. In a new chat open + → Apps/Connectors and enable the connector.

OpenAI has renamed this section before (Connectors → Apps/Plugins). If the label differs, search settings for "developer mode". On Business/Enterprise workspaces an admin must allow custom connectors first.

Authorization

On first use ChatGPT opens the service's sign-in window. Sign in and approve access.

How to connect llmtomd to Cursor

The server address to paste into Cursor settings:

https://llmtomd--support-0vd1.run.tools

Fastest: click Open in Cursor below and confirm the prompt.

Manually:

  1. In Cursor open Settings → Cursor Settings → MCP and click Add new global MCP server.
  2. Paste the JSON copied below into ~/.cursor/mcp.json (per project: .cursor/mcp.json in the repo root).
  3. Save the file. The server appears in the MCP list; authorize it there if asked.

Authorization

If the service needs sign-in, an authorize button appears next to the server in the MCP list.

Server tool list (7)

Raw names from tools/list. Only developers need these.

convert_urlConvert a web page or a publicly reachable file URL into clean, AI-ready Markdown. Call this when the user gives you a link (article, docs page, or a PDF/Office/image file URL) and wants it as Markdown, or wants to summarize or analyze its contents. Blocks until the conversion finishes (up to ~2 minutes) and returns the Markdown plus a document_id you can reuse with the other tools. Args: url: The web page or file URL to convert. crawl: If true, follow same-origin links and convert the whole site (bounded). ocr_langs: Optional Tesseract language hint for scanned files, e.g. "eng+fra".
list_documentsList the documents in the user's LLMtoMD library, most recent first. Call this to discover what the user has stored, or to find a document_id before reading, searching, or asking about a specific document. Pass `collection` (a project name or id) to list only that project's documents.
get_documentFetch a stored document's full Markdown by its document_id. Use the document_id returned by convert_url or list_documents. Call this when the user wants to read, summarize, or work with the full contents of a specific document.
search_documentsSemantic search across the user's LLMtoMD documents; returns the most relevant passages with similarity scores. Call this to locate where something is discussed across many documents — including specs, requirements/FRDs, and design docs when working on software. Pass document_id to search within a single document, or collection (a project name or id) to search only that project.
ask_documentsAnswer a question using the user's LLMtoMD documents (retrieval-augmented): finds the relevant passages and returns a cited answer with its sources. Call this when the user asks a question whose answer lives in their stored documents — including project requirements, specs, and API references when building software, so you rely on the authoritative spec rather than guessing. Pass document_id to scope the answer to a single document, or collection (a project name or id) to scope it to one project.
list_collectionsList the user's collections (projects) and how many documents each contains. A collection groups a project's documents (e.g. an FRD, specs, and design docs) so you can scope search_documents / ask_documents / list_documents to it with `collection`, or load it as context with the load_project prompt.
save_noteSave Markdown back into the user's LLMtoMD library so it persists and is searchable. Use this as write-back memory: record a decision, a spec you derived, a summary, or progress notes during a build, so the detail survives into later sessions instead of being lost when this conversation ends. Pass `collection` (a project name or id) to file the note under a project, or `append_to` (a note's document_id) to add to an existing note instead of creating a new one. Args: title: A short title for the note (used as its filename). content: The Markdown body to store. collection: Optional project to file the note under (name or id). append_to: Optional document_id of an existing note to append to.