openproject-mcp
Exposes the OpenProject REST API as MCP tools for project management, including creating and managing projects, work packages, relations, attachments, users, notifications, watchers, boards, and reference data.
README
openproject-mcp
An MCP server that exposes the OpenProject REST API (v3) as tools usable by Claude Desktop, Claude Code, Cursor, and any other MCP client.
Install
Before you start: you need an OpenProject API token. Get one from My account > Access tokens > API in your OpenProject instance.
Pick your MCP client and run one command:
Claude Code
claude mcp add openproject \
--env OPENPROJECT_BASE_URL=https://your-instance.openproject.com \
--env OPENPROJECT_API_KEY=your-token \
-- npx -y github:OliverRhyme/openproject-mcp
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS)
or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"openproject": {
"command": "npx",
"args": ["-y", "github:OliverRhyme/openproject-mcp"],
"env": {
"OPENPROJECT_BASE_URL": "https://your-instance.openproject.com",
"OPENPROJECT_API_KEY": "your-token"
}
}
}
}
Cursor
Add to .cursor/mcp.json in your project root:
{
"mcpServers": {
"openproject": {
"command": "npx",
"args": ["-y", "github:OliverRhyme/openproject-mcp"],
"env": {
"OPENPROJECT_BASE_URL": "https://your-instance.openproject.com",
"OPENPROJECT_API_KEY": "your-token"
}
}
}
}
Windsurf
Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"openproject": {
"command": "npx",
"args": ["-y", "github:OliverRhyme/openproject-mcp"],
"env": {
"OPENPROJECT_BASE_URL": "https://your-instance.openproject.com",
"OPENPROJECT_API_KEY": "your-token"
}
}
}
}
Any other MCP client (generic stdio)
Run this as the server process:
OPENPROJECT_BASE_URL=https://your-instance.openproject.com \
OPENPROJECT_API_KEY=your-token \
npx -y github:OliverRhyme/openproject-mcp
The server speaks JSON-RPC over stdio. No build step needed — npx fetches, installs, and runs it directly from GitHub.
Configuration
| Variable | Required | Default | Notes |
|---|---|---|---|
OPENPROJECT_BASE_URL |
yes | -- | e.g. https://community.openproject.org |
OPENPROJECT_API_KEY |
yes | -- | From My account > Access tokens > API |
OPENPROJECT_PAGE_SIZE |
no | 25 |
Default page size for list endpoints |
OPENPROJECT_TIMEOUT_MS |
no | 30000 |
HTTP request timeout in milliseconds |
Features
- Projects -- list, get, create, update, delete, count
- Work packages -- list, get, create, update (with
lockVersion), delete, activity/comment thread, inline file attachments on comments, count - Relations -- list, get, create, delete (blocks, precedes, relates, duplicates, etc.)
- Attachments -- list, get (with download), upload (with auto-embed in description or comment), delete
- Users -- current user, list, get
- Notifications -- list, get, mark read, mark all read
- Watchers -- list, add, remove watchers on work packages
- Boards -- list and get Kanban-style boards
- Reference data -- work package types, statuses, priorities, versions
- Raw passthrough -- call any GET endpoint under
/api/v3directly - Filter, sort, group, and paginate via OpenProject's native query syntax
- Output optimization -- field selection, description truncation,
hasMorepagination flag, lightweight count tools
Tools reference
All tools are prefixed op_ to avoid collisions with other MCP servers.
Projects
| Tool | Description |
|---|---|
op_list_projects |
List projects with filter/sort/paginate |
op_get_project |
Get project by id or identifier slug |
op_create_project |
Create a new project |
op_update_project |
Patch project fields |
op_delete_project |
Delete a project (destructive, async) |
op_count_projects |
Count projects matching filters |
Work packages
| Tool | Description |
|---|---|
op_list_work_packages |
List work packages, optionally scoped to a project |
op_get_work_package |
Get a single work package by id |
op_create_work_package |
Create a work package (requires subject, projectId, typeId) |
op_update_work_package |
Patch work package fields (requires lockVersion) |
op_delete_work_package |
Delete a work package (destructive) |
op_list_work_package_activities |
List comments and change history |
op_comment_work_package |
Add a comment, optionally attaching and embedding a file inline |
op_count_work_packages |
Count work packages matching filters |
Relations
| Tool | Description |
|---|---|
op_list_relations |
List relations on a work package (blocks, precedes, duplicates, etc.) |
op_get_relation |
Get a single relation by id |
op_create_relation |
Create a relation between two work packages |
op_delete_relation |
Delete a relation |
Attachments
| Tool | Description |
|---|---|
op_list_attachments |
List attachments on a work package |
op_get_attachment |
Get attachment metadata; optionally download to a local path |
op_upload_attachment |
Upload a local file; optionally embed in a comment or WP description |
op_delete_attachment |
Delete an attachment (destructive) |
Users
| Tool | Description |
|---|---|
op_current_user |
Get the user tied to the configured API key |
op_list_users |
List users with filter/sort/paginate |
op_get_user |
Get a user by id |
Notifications
| Tool | Description |
|---|---|
op_list_notifications |
List in-app notifications for the current user |
op_get_notification |
Get a single notification by id |
op_mark_notification_read |
Mark one notification as read |
op_mark_all_notifications_read |
Mark all notifications as read |
Watchers
| Tool | Description |
|---|---|
op_list_watchers |
List users watching a work package |
op_add_watcher |
Add a user as a watcher |
op_remove_watcher |
Remove a user from watching |
Boards
| Tool | Description |
|---|---|
op_list_boards |
List Kanban-style boards for a project |
op_get_board |
Get board details including column config |
Reference data
| Tool | Description |
|---|---|
op_list_types |
List work package types (Task, Bug, Feature, etc.) |
op_list_statuses |
List all work package statuses |
op_list_priorities |
List all priorities |
op_list_versions |
List versions/milestones, optionally per project |
op_api_passthrough |
Raw GET against any /api/v3 path (escape hatch) |
Usage patterns
Filter syntax
List endpoints accept OpenProject's structured filter format:
{
"projectIdOrIdentifier": "web",
"filters": [
{ "field": "status_id", "operator": "o", "values": null },
{ "field": "assignee", "operator": "=", "values": ["42"] },
{ "field": "type", "operator": "=", "values": ["1"] }
],
"sortBy": [["updatedAt", "desc"]],
"pageSize": 50
}
Common operators:
| Operator | Meaning |
|---|---|
= |
Equals |
! |
Not equals |
~ |
Contains (substring) |
o |
Open statuses |
c |
Closed statuses |
>= |
Greater or equal |
<= |
Less or equal |
* |
Any (not empty) |
!* |
None (empty) |
Filter values are always strings, even for numeric ids: "values": ["42"], not [42].
Output optimization
List tools return summarized output by default. Several options keep responses small:
fields -- Return only specific fields per element:
// op_list_work_packages
{ "fields": ["id", "subject", "status"] }
// -> elements contain only { id, subject, status }
raw -- Get the full HAL+JSON document from OpenProject:
// op_get_work_package
{ "id": 17, "raw": true }
Count tools -- When you only need a number, use op_count_work_packages or op_count_projects instead of listing:
// op_count_work_packages
{ "projectIdOrIdentifier": "web", "filters": [{ "field": "status_id", "operator": "o", "values": null }] }
// -> { "total": 42 }
Pagination -- All list responses include hasMore: true|false so you know if there are more pages. Max pageSize is 100; default is 25.
Truncation -- Project descriptions are truncated to 200 chars in list mode. Activity comments are truncated to 500 chars by default; pass full: true to op_list_work_package_activities for complete text.
Updating work packages (lockVersion)
OpenProject uses optimistic locking. You must pass the current lockVersion
when updating a work package -- fetch it first with op_get_work_package:
// 1. Get current state
// op_get_work_package { "id": 17 }
// -> { ..., "lockVersion": 4 }
// 2. Update with lockVersion
// op_update_work_package
{
"id": 17,
"lockVersion": 4,
"statusId": 7,
"percentageDone": 50
}
Uploading and embedding attachments
Upload a file and auto-embed it as an image in a comment:
// op_upload_attachment
{
"workPackageId": 17,
"filePath": "/path/to/screenshot.png",
"embedIn": "comment",
"embedText": "Here's the updated design:"
}
Or attach a file when commenting:
// op_comment_work_package
{
"id": 17,
"comment": "Fixed in latest build, see attached screenshot.",
"attachFilePath": "/path/to/screenshot.png"
}
Creating relations
Link work packages with dependency or reference relations:
// op_create_relation
{
"fromId": 17,
"toId": 23,
"type": "blocks"
}
Relation types: relates, duplicates, blocks, precedes, follows,
includes, partOf, requires.
Local development
git clone https://github.com/OliverRhyme/openproject-mcp.git
cd openproject-mcp
npm install
npm run dev
| Command | Purpose |
|---|---|
npm install |
Install dependencies |
npm run build |
Compile TypeScript to dist/ |
npm start |
Run the compiled server (requires build) |
npm run dev |
Watch-mode server via tsx (no build) |
npm test |
Run test suite |
npm run test:watch |
Run tests in watch mode |
npm run typecheck |
Type-check without emitting |
Smoke test
Verify the server boots and lists tools without a real OpenProject instance:
OPENPROJECT_BASE_URL=https://example.openproject.com OPENPROJECT_API_KEY=fake \
node dist/index.js <<'EOF'
{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"smoke","version":"0.0.0"}}}
{"jsonrpc":"2.0","method":"notifications/initialized"}
{"jsonrpc":"2.0","id":2,"method":"tools/list"}
EOF
License
MIT
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.