Writespace
Persistent docs and memory for AI agents. Writespace is a collaborative markdown editor with a built-in MCP server — your model reads, writes, organizes, and searches a shared workspace while humans edit the same docs live. Drop the ranked full-text search straight in as RAG retrieval.
README
Writespace MCP Server
A remote, hosted Model Context Protocol server for Writespace — a real-time collaborative document editor. Give Claude, ChatGPT, Gemini, Cursor, or any MCP-speaking agent a shared docs workspace it can read, write, organise and search alongside you.
Endpoint: https://app.writespace.io/mcp (HTTP transport, Bearer auth)
Free tier: yes — up to 10 documents, no card required → sign up
What it does
Writespace is a markdown-first doc editor for humans with an MCP server built in for agents. Every action you can take in the app is exposed as a tool your model can call:
- Read — walk the workspace tree, fetch any doc as lossless JSON or markdown, run ranked full-text search (drop it straight into your agent as RAG retrieval)
- Write — create docs, append content, replace a single section by its heading, shallow-merge structured metadata
- Organise — create/rename/move folders, move docs, query docs by metadata (
{ status: "accepted" })
Agent writes broadcast over Realtime, so open editors reload live. Humans and agents work on the same canvas — no export step, no format negotiation.
Quick start
Three steps, two minutes:
- Get a token — sign in at app.writespace.io, open Settings → API tokens, create a token (shown once, revocable any time).
- Endpoint — one hosted endpoint for all accounts:
https://app.writespace.io/mcp - Configure your client — snippets below.
Claude Desktop
~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"writespace": {
"type": "http",
"url": "https://app.writespace.io/mcp",
"headers": {
"Authorization": "Bearer ws_pat_…"
}
}
}
}
Claude Code (CLI)
claude mcp add --transport http writespace \
https://app.writespace.io/mcp \
--header "Authorization: Bearer ws_pat_…"
Cursor
Settings → MCP → Add server
Type: HTTP
URL: https://app.writespace.io/mcp
Header: Authorization: Bearer ws_pat_…
Gemini CLI
~/.gemini/settings.json
{
"mcpServers": {
"writespace": {
"httpUrl": "https://app.writespace.io/mcp",
"headers": {
"Authorization": "Bearer ws_pat_…"
}
}
}
}
Any other spec-compliant MCP client works the same way: HTTP transport + Authorization: Bearer header.
Tools
Every tool except list_workspaces takes a workspace_id — call list_workspaces first to discover them.
| Tool | Description |
|---|---|
list_workspaces |
List every workspace you belong to |
list_folders |
List folders, optionally nested under a parent |
list_docs |
List documents, optionally scoped to a folder; returns metadata per row |
get_doc |
Fetch a doc by ID — canonical Tiptap JSON + markdown rendering + metadata |
search |
Ranked full-text search with <<match>> snippets |
query_docs |
Find docs whose metadata contains a JSON object (Postgres JSONB @>) |
create_doc |
Create a doc from Tiptap JSON or markdown |
update_doc |
Update title, content, or metadata (shallow-merge; null deletes a key) |
upsert_doc |
Create-or-update by title within an optional folder — idempotent |
append_to_doc |
Append markdown / Tiptap JSON to an existing doc |
replace_section |
Replace a section's body by its heading text — surgical edits |
delete_doc |
Delete a doc by ID |
move_doc |
Move a doc to another folder (folder_id: null → root) |
create_folder |
Create a folder, optionally nested |
rename_folder |
Rename a folder — IDs stay stable, links don't break |
delete_folder |
Delete a folder and its contents |
Full reference with format notes and troubleshooting: writespace.io/connect
Design notes
- Lossless content, two formats — reads return both
content(Tiptap JSON) andcontent_markdown; writes accept either. Markdown round-trips for everything the editor supports, including tables, task lists, and mermaid diagrams. - Lean write responses — write tools return only metadata by default, saving 80–95% of the payload on long docs. Pass
include_content: trueto echo the body. - Structured metadata — every doc carries a JSONB
metadatafield. Treat it as queryable frontmatter: write it, shallow-merge it, query by containment. - Stable cross-doc links — reference docs as
writespace://doc/<id>; IDs survive renames and moves. - Live reload — MCP writes notify open editors over Realtime; they refetch in the same tab.
Use cases
- MCP docs server — a docs backend your agents treat as a tool
- Shared docs for Claude — persistent memory and notes for Claude
- Shared docs for ChatGPT — same workspace, different model
- MCP knowledge base — searchable team knowledge any agent can query
Auth & security
- Per-agent personal access tokens, issued and revoked from Settings — scoped to your account, easy to rotate
- Tokens are shown once at creation
- Standard
Authorization: Bearer ws_pat_…header; spec-compliant HTTP transport
Links
- Website: writespace.io
- App: app.writespace.io
- Setup & full tool reference: writespace.io/connect
- Contact: hello@writespace.io
© 2026 Writespace · Pro is $5/user/month, free tier available.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.