Targetprocess MCP Server

Targetprocess MCP Server

The Model Context Protocol (MCP) is a standard that enables AI assistants to interact with external tools and services through a unified interface. MCP servers provide these capabilities by exposing tools and resources that AI assistants can use to accomplish tasks.

This MCP server provides tools for interacting with Targetprocess, a project management and agile planning platform. It enables AI assistants to:

• Search and retrieve Targetprocess entities (User Stories, Bugs, Tasks, Features, etc.)
• Create and update entities with proper validation
• Query entities with complex filters and includes
• Inspect and discover the Targetprocess data model
• Handle authentication and API interactions safely

Key Features

• Data Model Discovery: Explore and understand complex Targetprocess implementations
• Powerful Querying: Use complex filters and includes to retrieve exactly the data you need
• Entity Management: Create and update entities with proper validation
• Relationship Exploration: Understand how different entities relate to each other
• Error Handling: Robust error handling with retries and informative messages
• Documentation Integration: Built-in access to Targetprocess documentation

Use Cases

This MCP server is particularly valuable in corporate settings where Targetprocess might handle millions of records with complex schemas and data models. Common use cases include:

• Data Model Discovery: Map and understand complex Targetprocess implementations
• Enterprise Analytics: Extract and analyze data across millions of records
• Cross-System Integration: Use as a bridge between Targetprocess and other systems
• Custom Reporting: Build specialized reports not available in the standard UI
• Batch Operations: Manage large-scale changes across many entities
• Schema Exploration: Discover custom fields and relationships in complex implementations

For detailed examples and implementation guides, see USECASES.md.

Getting Started

Clone the repository recursively to include the documentation search tool:

Development Resources

Documentation Search

This repository includes a documentation scraper/searcher for Targetprocess developer documentation as a submodule. You can use it to quickly search through Targetprocess's documentation:

The search tool is located in resources/target-process-docs. We use pushd/popd commands here because:

1. The tool requires access to its database files using relative paths
2. pushd saves your current directory location
3. Temporarily changes to the tool's directory to run the command
4. popd automatically returns you to your previous location This approach lets you run searches from any directory while ensuring the tool can find its database files.

This tool provides a powerful way to search through Targetprocess's developer documentation locally. The search results include relevant documentation sections with context, making it easier to find specific API details or implementation guidance.

CI/CD Pipeline

The project uses GitHub Actions for automated builds:

• Pushes to main branch trigger new container builds
• Version tags (v*..) create versioned releases
• Images are published to GitHub Container Registry

You can use the published image:

Environment Variables

• TP_DOMAIN: Your Targetprocess domain (e.g., company.tpondemand.com)
• TP_USERNAME: Your Targetprocess username
• TP_PASSWORD: Your Targetprocess password

Local Development with Docker

For local development and testing, use the provided scripts:

1. Build the local image:

   Note: The build script uses Docker's quiet mode by default to minimize log output. This is intentional to reduce AI token consumption when interacting with tools like Cline that process the build output. In quiet mode, the full build log is saved to /tmp/apptio-target-process-mcp/docker-build.log. Use --verbose flag to see build output directly in the terminal.

2. Run the local image:

3. Configure Cline: Edit ~/.config/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json:

Local Development without Docker

Prerequisites

• Node.js 20 or later
• npm

Setup

1. Clone the repository recursively:

Note: The --recursive flag is required to also clone the documentation search tool submodule.

2. Install dependencies:

3. Copy the example config:

4. Edit config/targetprocess.json with your Targetprocess credentials.

Building

Running

API Capabilities

For detailed examples and common use cases, see USECASES.md.

The MCP server provides the following tools for interacting with Targetprocess:

search_entities

Search for Targetprocess entities (UserStory, Bug, Task, Feature) with filtering and includes.

get_entity

Get detailed information about a specific entity.

create_entity

Create a new entity in Targetprocess.

update_entity

Update an existing entity.

inspect_object

Inspect Targetprocess objects and properties through the API.

Performance Considerations

When working with large Targetprocess instances that may contain millions of records:

1. Use Specific Queries: Always use the most specific query possible to limit result sets
2. Limit Result Size: Use the take parameter to limit the number of results returned
3. Include Only Necessary Data: Only include related data that you actually need
4. Consider Pagination: For large result sets, implement pagination in your application
5. Batch Operations: For bulk operations, consider batching requests to avoid overloading the API

LLM Integration

This MCP server can be used with various AI assistants that support the Model Context Protocol:

• Cline - A CLI-based AI assistant
• Claude Desktop - Anthropic's desktop application
• Goose - A local AI assistant

For configuration and setup instructions, see llms-install.md.

Configuration

The server can be configured either through environment variables or a JSON config file.

Config File Format

License

MIT