ProofHub MCP Server
README.md
# ๐ ProofHub MCP Server
[](https://github.com/MustafaPatharia/proofhub-mcp)
[](LICENSE)
[](https://modelcontextprotocol.io)
A **Model Context Protocol (MCP)** server that bridges ProofHub's project management capabilities with AI assistants like Claude, GitHub Copilot, and Cursor. Paste a ProofHub task URL in your AI chat, and the agent automatically fetches task details, comments, and bug tracker links to help you complete development tasks faster.
## ๐ฏ What This Solves
As a developer working with ProofHub for task management, you normally have to:
1. Open ProofHub in a browser
2. Navigate to the task
3. Read the description and all comments
4. Copy relevant bug tracker links (Jira, Linear, GitHub Issues, etc.)
5. Context-switch back to your code editor
**With this MCP server**, simply paste the ProofHub URL into your AI assistant, and it automatically:
- โ
Fetches the complete task description
- โ
Retrieves all comments and history
- โ
Extracts bug tracker links from anywhere in the task
- โ
Provides structured context for the AI to help you code
## โก Features
- **๐ Secure**: API key stored in environment variables, never in code
- **โฑ๏ธ Rate Limit Handling**: Automatically respects ProofHub's 25 req/10s limit with retry logic
- **๐งน HTML Cleanup**: Strips HTML tags from descriptions/comments for clean AI consumption
- **๐ Smart Link Extraction**: Regex patterns detect bug tracker URLs across all text fields
- **๐ฏ One-Shot Tool**: `get_task_with_bug_links` fetches everything in a single call
- **๐ก๏ธ Error Handling**: Graceful error messages with API status codes
## ๐ Quick Start
### 1. Installation
```bash
npm install
```
### 2. Configuration
Create a `.env` file (use `.env.example` as template):
```env
PROOFHUB_API_KEY=your_api_key_here
PROOFHUB_API_BASE=https://company.proofhub.com/api/v3
```
> **Getting your API key:** Log in to ProofHub โ Account Settings โ Apps & Integrations โ API โ Generate new API key
### 3. Connect to Your AI Assistant
#### ๐ค Claude Desktop
Add to your `claude_desktop_config.json`:
```json
{
"mcpServers": {
"proofhub": {
"command": "node",
"args": ["/absolute/path/to/proofhub-mcp/index.js"],
"env": {
"PROOFHUB_API_KEY": "your_key_here"
}
}
}
}
```
**Config location:**
- macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
- Windows: `%APPDATA%\Claude\claude_desktop_config.json`
#### ๐ป VS Code Copilot
Add to `.vscode/mcp.json` in your workspace or user settings:
```json
{
"servers": {
"proofhub": {
"type": "stdio",
"command": "node",
"args": ["/absolute/path/to/proofhub-mcp/index.js"],
"env": {
"PROOFHUB_API_KEY": "your_key_here"
}
}
}
}
```
## ๐ Available Tools
| Tool | Description | Use Case |
|------|-------------|----------|
| **`proofhub_parse_url`** | Extracts project/list/task IDs from a ProofHub URL | Parse URL before calling other tools |
| **`proofhub_get_task`** | Fetches complete task details (title, description, stage, custom fields, assignees) | Get full task context |
| **`proofhub_get_comments`** | Retrieves all task comments with timestamps | Review team discussions |
| **`proofhub_get_task_with_bug_links`** | **โญ One-shot**: Fetches task + comments + auto-extracts bug tracker URLs | **Most useful** - get everything in one call |
| **`proofhub_create_comment`** | Posts a new comment on a task | Update task from your AI chat |
| **`proofhub_get_task_history`** | Fetches activity/stage change history | Track task evolution |
## ๐ก Usage Examples
### Example 1: Fetch Task Context
Simply paste a ProofHub URL in your AI chat:
```
Get context for this task:
https://kpi.proofhub.com/bappswift/#app/todos/project-7189443252/list-270280503800/task-514774338823
```
The AI will call `proofhub_get_task_with_bug_links` and return:
- Task title and description
- All comments (cleaned from HTML)
- List of bug tracker links found
- Stage, assignees, and custom fields
### Example 2: Development Workflow
```
I need to work on this ProofHub task:
https://kpi.proofhub.com/.../task-514774338823
Fetch the task, find any Jira links, and help me understand what needs to be built.
```
The AI will:
1. Fetch the complete task context
2. Extract Jira/GitHub/Linear links
3. Use your bug tracker MCP to fetch issue details
4. Provide a comprehensive development plan
### Example 3: Update Task
```
Add a comment to task 514774338823 in project 7189443252, list 270280503800:
"Completed the API endpoint implementation. Ready for review."
```
## ๐ค Contributing
Contributions welcome! Fork the repo, make your changes, and submit a pull request.
## References
- [ProofHub API v3](https://github.com/ProofHub/api_v3)
TDQS
A3.5/5.0
Scored across 6 tools
Disambiguation5/5
Each tool has a clearly distinct purpose: creating comments, fetching comments, getting task details, getting task history, a composite bug-link extractor, and a URL parser. No overlap.
Naming Consistency5/5
All tools use the 'proofhub_' prefix followed by consistent snake_case verb_noun patterns (e.g., create_comment, get_task, parse_url). Naming is uniform and predictable.
Tool Count5/5
With 6 tools, the server is well-scoped for interacting with ProofHub tasks. Each tool serves a clear purpose without bloat or deficiency.
Completeness3/5
The tool set covers creation, reading, and composite extraction, but lacks update, delete, and list operations for tasks and comments. This creates notable gaps for full lifecycle management.