targetprocess-mcp
MCP server for integrating TargetProcess with Claude Code, enabling natural language queries for projects, user stories, bugs, features, and sprints.
README
TargetProcess MCP
Model Context Protocol (MCP) server for integrating TargetProcess with Claude Code and other MCP clients.
Features
- Secure credential storage - API tokens stored in macOS Keychain
- VPN protection - Optional requirement for VPN connection
- Async API client - Built on httpx for efficient HTTP requests
- Type-safe models - Dataclass models for TargetProcess entities
- Self-configuring - Configure directly from Claude Code
Requirements
- Python 3.10+
- macOS (for Keychain integration)
- TargetProcess account with API access
- (Optional) VPN for secure access
Quick Start
1. Install
# Install uv if you don't have it
curl -LsSf https://astral.sh/uv/install.sh | sh
# Clone and install
cd targetprocess-mcp
uv sync
2. Add to Claude Code
# Using CLI (recommended)
claude mcp add targetprocess --transport stdio --scope user -- python -m targetprocess_mcp.server
Or manually add to ~/.claude.json:
{
"mcpServers": {
"targetprocess": {
"command": "python",
"args": ["-m", "targetprocess_mcp.server"]
}
}
}
3. Configure
The first time you use a tool, you'll be prompted to configure. Or run the configure tool directly:
configure(url="https://yourcompany.tpondemand.com", token="your-api-token")
This stores:
- URL:
~/.config/targetprocess-mcp/config.toml - Token: macOS Keychain (encrypted)
Available Tools
| Tool | Description |
|---|---|
configure |
Configure URL, token, and VPN settings |
get_status |
Check if MCP is configured and connected |
get_projects |
List all projects (id + name for reference) |
search |
Search user stories, bugs, or features by name |
get_user_stories |
Query user stories with filters |
get_bugs |
Query bugs with filters |
get_features |
Query features with filters |
get_sprints |
Query sprints/releases |
Tool Details
configure
Configure TargetProcess MCP with your credentials. Run this first to set up access.
configure(
url: str, # TargetProcess URL
token: str, # API token
vpn_required: bool = false, # Require VPN connection
vpn_check_hosts: str = None # Comma-separated hosts to check
)
Example:
configure(url="https://mycompany.tpondemand.com", token="abc123")
get_status
Check if TargetProcess MCP is configured and connected.
get_projects
Get all projects for quick reference. Use the returned IDs for filtering other queries.
# Returns: [{"id": 1, "name": "My Project"}, ...]
get_projects(take: int = 100)
search
Search entities by name across user stories, bugs, and features.
search(
query: str, # Search term
entity_type: str = None, # UserStory, Bug, Feature, or Task
project_id: int = None, # Filter by project
take: int = 20 # Max results
)
get_user_stories
Query user stories with optional filters.
get_user_stories(
project_id: int = None, # Filter by project
feature_id: int = None, # Filter by feature
assignee_id: int = None, # Filter by assignee
state: str = None, # e.g., "Open", "In Progress", "Done"
take: int = 100
)
get_bugs
Query bugs with optional filters.
get_bugs(
project_id: int = None, # Filter by project
assignee_id: int = None, # Filter by assignee
state: str = None, # e.g., "Open", "Resolved", "Done"
severity: str = None, # e.g., "Critical", "Major", "Minor"
take: int = 100
)
get_features
Query features with optional filters.
get_features(
project_id: int = None, # Filter by project
state: str = None, # e.g., "Proposed", "In Progress"
take: int = 100
)
get_sprints
Query releases/sprints.
get_sprints(
project_id: int = None, # Filter by project
take: int = 50
)
Usage Examples
Find all open bugs in a project
Get all open bugs in project 5
Search for a user story
Search for user stories named "login"
Get features for a project
Get all features for project 3
Find sprints for a specific project
Show me the sprints for project "MyApp"
Natural Language Prompts
Here are examples of how to interact with TargetProcess using natural language:
Getting Started
Configure TargetProcess with my account (url)
Show me my TargetProcess status
Projects & Search
What projects do I have access to?
Find projects named "Mobile"
Search for user stories about login
Find bugs tagged as critical
User Stories
Show my open user stories
Get all user stories for project 5
List user stories assigned to me that are in progress
Show me user stories for feature 123
Bugs
Show me all open bugs
What critical bugs exist in project 3?
List bugs assigned to me
Show bugs in QA state for review
Features
What features are in progress?
Show completed features for project "Website Redesign"
Sprints/Releases
What sprints are coming up?
Show me the current sprint for project 5
List all releases for project "Mobile App"
Security
Credential Storage
- API Token: Stored in macOS Keychain (encrypted by OS)
- URL & Settings: Stored in
~/.config/targetprocess-mcp/config.toml
VPN Protection (Optional)
Configure VPN protection when running configure:
vpn_required: true
vpn_check_hosts: internal-api.company.com
When enabled, the MCP will check connectivity to the specified host(s) before making any API calls.
Configuration
Re-configure
Run the configure tool anytime to update your settings:
configure(url="https://yourcompany.tpondemand.com", token="new-token")
configure(url="https://yourcompany.tpondemand.com", token="token", vpn_required=true, vpn_check_hosts="host1,host2")
Environment Variables
| Variable | Description |
|---|---|
TARGETPROCESS_URL |
Override TargetProcess URL |
TARGETPROCESS_TOKEN |
Override API token (not recommended) |
TARGETPROCESS_VPN_REQUIRED |
Set to "true" to require VPN |
Manual Configuration
The configuration file is stored at:
~/.config/targetprocess-mcp/config.toml
Example:
URL = "https://yourcompany.tpondemand.com"
VPN_REQUIRED = true
VPN_CHECK_HOSTS = ["internal-api.company.com"]
Troubleshooting
"TargetProcess not configured" error
Run the configure tool:
configure(url="https://yourcompany.tpondemand.com", token="your-api-token")
"VPN connection required" error
Ensure you're connected to your VPN. If you configured VPN protection, the MCP will not work outside your network.
Claude Code not recognizing the MCP
Try restarting Claude Code or running:
claude mcp list
To verify the server is registered.
macOS Keychain issues
If you encounter Keychain errors, you can manually manage the token:
# Add token
security add-generic-password -s targetprocess-mcp -a api-token -w YOUR_TOKEN
# Delete token
security delete-generic-password -s targetprocess-mcp -a api-token
Development
Project Structure
targetprocess_mcp/
├── __init__.py # Package metadata
├── config.py # Configuration & Keychain
├── client.py # TargetProcess API client
├── server.py # FastMCP server & tools
├── models/ # Dataclass models
│ ├── __init__.py
│ ├── project.py
│ ├── user_story.py
│ └── ...
├── pyproject.toml # Package configuration
└── .pre-commit-config.yaml
Running Tests
uv run pytest tests/
Code Quality
# Run ruff linter
uv run ruff check targetprocess_mcp/
# Run mypy type checker
uv run mypy targetprocess_mcp/
Pre-commit Hook
Install the pre-commit hook to run type checks and tests before each commit:
# Install pre-commit
pip install pre-commit
# Install hooks
pre-commit install
# Run manually anytime
pre-commit run --all-files
The hook runs:
ruff- Linting and formattingmypy- Type checkingpytest- Tests
Commit is blocked if any check fails.
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。