Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@ActiveCollab MCP Servershow my open tasks and recent notifications"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
ActiveCollab MCP Server
Local MCP (Model Context Protocol) server for ActiveCollab API integration. Allows Claude Code to access ActiveCollab data without exposing API tokens to external services.
Architecture
Requirements
Python 3.8+
pip
Setup
1. Ensure Python is available
On some Linux distributions, python command may not exist (only python3). Fix with:
2. Install dependencies
3. Generate API token
Run the token generator script:
The script will prompt you for:
ActiveCollab URL - Your self-hosted instance URL (e.g.,
https://collab.yourcompany.com)Email - Your ActiveCollab login email
Password - Your ActiveCollab password (hidden input)
How it works:
Makes a POST request to
/api/v1/issue-tokenwith your credentialsActiveCollab returns a permanent API token for the "Claude MCP Server" client
Token is saved to
config.json
Alternative: Manually create config.json:
4. Configure MCP server
5. Restart Claude Code
Available Tools
Tool | Parameters | Description |
| - | List all projects |
|
| Get project details |
|
| List tasks for a project |
|
| Get task details |
| - | Get tasks assigned to current user |
|
| Get comments on a task |
|
| Get time records for a task |
|
| Get all time records for a project |
| - | Get recent notifications for current user |
|
| Get subtasks for a task |
|
| Get tasks with overdue deadlines |
|
| Get open tasks sorted by due date |
| - | Get all global task labels |
|
| Get tasks with specific label |
|
| Get all task lists in a project |
|
| Get tasks in a specific list |
| - | Get all team members (not clients) |
|
| Get tasks assigned to a user |
Usage Examples
Project Structure
Dependencies
Python 3.8+ (ensure
pythoncommand is available, see Setup step 1)mcp>=1.0.0- Model Context Protocol SDKhttpx>=0.25.0- HTTP client (optional, using urllib)
Implementation Status
Phase 1: Basic infrastructure (config, client, auth)
Phase 2: MCP server skeleton
Phase 3: Projects module (list, get)
Phase 4: Tasks module (list, get, filter by status)
Phase 5: Time tracking module (read only)
Phase 6: Comments module (read only)
Phase 7: Notifications module (read only)
Phase 8: Subtasks module (read only)
Phase 9: Deadlines module (overdue tasks, tasks with deadlines)
Phase 10: Labels module (list labels, filter tasks by label)
Phase 11: Task lists module (list task lists, filter tasks by list)
Phase 12: Users module (list users, filter tasks by assignee)
Phase 13: Write operations (create/update tasks, add comments, log time)