OpenProjectSAFE
A Model Context Protocol (MCP) server that provides seamless integration with OpenProject API v3.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://mseep.ai/app/andyeverything-openproject-mcp-server)
⚠️ This is an early-stage project. Do not use it productively – contributions welcome!
A Model Context Protocol (MCP) server that provides seamless integration with OpenProject API v3. This server enables LLM applications to interact with OpenProject for project management, work package tracking, and task creation.
Features
- 🔌 Full OpenProject API v3 Integration
- 📋 Project Management: List and filter projects
- 📝 Work Package Management: Create, list, and filter work packages
- 🏷️ Type Management: List available work package types
- 🔐 Secure Authentication: API key-based authentication
- 🌐 Proxy Support: Optional HTTP proxy configuration
- 🚀 Async Operations: Built with modern async/await patterns
- 📊 Comprehensive Logging: Configurable logging levels
Prerequisites
- Python 3.10 or higher
- uv (fast Python package manager)
- An OpenProject instance (cloud or self-hosted)
- OpenProject API key (generated from your user profile)
Installation
1. Install uv (if not already installed)
macOS/Linux:
curl -LsSf https://astral.sh/uv/install.sh | sh
Windows:
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
Alternative (using pip):
pip install uv
2. Clone and Setup the Project
git clone https://github.com/yourusername/openproject-mcp.git cd openproject-mcp
3. Create Virtual Environment and Install Dependencies
# Create virtual environment and install dependencies in one command uv sync
Alternative (manual steps):
# C
3b42657d35dbOBSERVED · 2026-10-07Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add openproject-mcp-server --env MCP_API_KEYS=${MCP_API_KEYS} --env OPENPROJECT_API_KEY=${OPENPROJECT_API_KEY} -- uvx openproject-mcp-server{
"mcpServers": {
"openproject-mcp-server": {
"command": "uvx",
"args": [
"openproject-mcp-server"
],
"env": {
"MCP_API_KEYS": "${MCP_API_KEYS}",
"OPENPROJECT_API_KEY": "${OPENPROJECT_API_KEY}"
}
}
}
}Exposed tools (46)
25 read · 14 write · 7 destructive. Blast radius: 7 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
check_permissions | read | Check current user permissions and capabilities |
create_membership | write | Create a new project membership |
create_news | write | Create a new news entry for a project. |
create_project | write | Create a new project |
create_time_entry | write | Create a new time entry |
create_version | write | Create a new project version/milestone |
create_work_package | write | Create a new work package with optional date fields |
create_work_package_relation | write | Create a relationship between work packages |
delete_membership | destructive | Delete a membership |
delete_news | destructive | Delete a news entry permanently. |
delete_project | destructive | Delete a project |
delete_time_entry | destructive | Delete a time entry |
delete_work_package | destructive | Delete a work package |
delete_work_package_relation | destructive | Delete a work package relation |
get_membership | read | Get detailed information about a specific membership |
get_news | read | Get detailed information about a specific news entry. |
get_project | read | Get detailed information about a specific project |
get_role | read | Get detailed information about a specific role |
get_user | read | Get detailed information about a specific user |
get_work_package | read | Get detailed information about a specific work package |
get_work_package_relation | read | Get detailed information about a specific work package relation |
list_memberships | read | List project memberships |
list_news | read | List news entries with filtering and pagination. |
list_priorities | read | List available work package priorities |
list_project_members | read | List all members of a specific project |
list_projects | read | List all OpenProject projects |
list_roles | read | List all available roles |
list_statuses | read | List available work package statuses |
list_time_entries | read | List time entries |
list_time_entry_activities | read | List available time entry activities |
list_types | read | List available work package types |
list_user_projects | read | List all projects a specific user is assigned to |
list_users | read | List all users |
list_versions | read | List project versions/milestones |
list_work_package_children | read | List all child work packages of a parent |
list_work_package_relations | read | List work package relations with optional filtering |
list_work_packages | read | List work packages with optional pagination |
remove_work_package_parent | destructive | Remove parent relationship from a work package (make it top-level) |
set_work_package_parent | write | Set a parent for a work package (create parent-child relationship) |
test_connection | read | Test the connection to the OpenProject API |
update_membership | write | Update an existing membership |
update_news | write | Update an existing news entry. |
update_project | write | Update an existing project |
update_time_entry | write | Update an existing time entry |
update_work_package | write | Update an existing work package including dates |
update_work_package_relation | write | Update an existing work package relation |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (5 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (8)
delete_membership, delete_news, delete_project, delete_time_entry, delete_work_package, delete_work_package_relation, remove_work_package_parent
Agile_Scrum_Weekly_Report_Template.docx
.fastmcp.yaml
mcp, aiohttp, python-dotenv, certifi
You are a senior API documenter with expertise in creating world-class API documentation. Your focus spans OpenAPI specification writing, interactive documentation portals, code example generation, an
curl -LsSf https://astral.sh/uv/install.sh | sh
curl -LsSf https://astral.sh/uv/install.sh | sh
Gates applied: no_behavioural_pass, no_license.
3b42657d35dbfull audit observations/trust-audit/mcp-server/andyeverything__openproject.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-07 | 3b42657d35db | SAFE | B | 89 | first audit |
Questions
What is the OpenProject MCP server?
A Model Context Protocol (MCP) server that provides seamless integration with OpenProject API v3.
What tools does OpenProject expose?
46 in total: 25 read-only, 14 that write, and 7 that can delete or overwrite (delete_membership, delete_news, delete_project, delete_time_entry, delete_work_package). Every one is listed on this page with its risk.
Is OpenProject safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 7 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does OpenProject need?
It reads MCP_API_KEYS and OPENPROJECT_API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does OpenProject run?
It speaks sse and stdio, so it runs as a local process your client starts. It is published on PyPI as openproject-mcp-server.
How current is this page?
The grade is for one exact copy of the source (3b42657d35db), read on 2026-10-07. The repository is watched and re-audited when it changes.