🧠 Targetprocess Model Context Protocol (MCP) Endpoint

Access Targetprocess Data via AI Assistants

This deployment serves as an implementation of the MCP specification, specifically tailored for seamless integration with the Targetprocess enterprise project management ecosystem. It transforms conversational input into structured API calls.

Empower your conversational agents to become proficient Targetprocess operators—manage backlogs, advance workflow states, and retrieve precise metrics through dialogue.

Deployment Pathways

Select the method that best suits your operational environment:

💻 Local Execution via NPX

For immediate testing or evaluation without persistent installation.

bash

Requires API Key Authentication (Recommended)

TP_DOMAIN=your-corp-instance.tpondemand.com TP_API_KEY=your-secure-token \ npx -y https://github.com/aaronsb/apptio-target-process-mcp.git

Or use Basic Authentication with role specialization

TP_DOMAIN=your-corp-instance.tpondemand.com TP_USERNAME=user@domain.com TP_PASSWORD=secretpassword \ TP_USER_ROLE=developer TP_USER_ID=1001 MCP_STRICT_MODE=true \ npx -y https://github.com/aaronsb/apptio-target-process-mcp.git

🐳 Containerized Operation with Docker

For robust, isolated deployment.

bash

Standard setup using environment variables for credentials

docker run -i --rm \ -e TP_DOMAIN=your-corp-instance.tpondemand.com \ -e TP_API_KEY=your-secure-token \ ghcr.io/aaronsb/apptio-target-process-mcp

Advanced setup specifying user context and strict JSON output

docker run -i --rm \ -e TP_DOMAIN=your-corp-instance.tpondemand.com \ -e TP_USERNAME=user@domain.com \ -e TP_PASSWORD=secretpassword \ -e TP_USER_ROLE=tester \ -e MCP_STRICT_MODE=true \ ghcr.io/aaronsb/apptio-target-process-mcp

🖥️ Integrated with Claude Desktop

For native experience within the Anthropic desktop client.

{ "mcpServers": { "targetprocess": { "command": "npx", "args": ["-y", "https://github.com/aaronsb/apptio-target-process-mcp.git"], "env": { "TP_DOMAIN": "your-corp-instance.tpondemand.com", "TP_API_KEY": "your-secure-token" } } } }

Credential Management and Configuration

Authentication Mechanisms

1. Token-Based Access (Preferred):

   • Generate an Access Token within Targetprocess (Settings → Access Tokens).
   • Assign this token to the TP_API_KEY environment variable.

2. Traditional Basic Authentication:

   • Set TP_USERNAME (login identifier).
   • Set TP_PASSWORD (corresponding credential).

⚠️ Security Precaution: Credentials must be injected via environment variables or secure configuration files (.env managed by .gitignore); never hardcode them in source code repositories.

Exposed Functionality (Core Tools)

The gateway exposes the following semantic operations to the connected AI model:

Tool Identifier	Functionality Description
search_entities	Execute complex searches incorporating filtering, required attribute inclusion, and ordering.
get_entity	Fetch exhaustive details for a single specified Targetprocess item.
create_entity	Provision new work items (stories, bugs, tasks) with integrated schema validation.
update_entity	Modify attributes of existing items in the system.
inspect_object	Introspect the available data structures and property schemas within the Targetprocess instance.

Contextual Role Augmentation

By defining a TP_USER_ROLE, the AI gains access to specialized actions tailored for specific user personas:

Role Designation	Supplemental Toolset
developer	Tools for personal task triage (show_my_tasks), work initiation (start_working_on), closure (complete_task), defect review (show_my_bugs), and time logging (log_time).
project-manager	Advanced portfolio monitoring and team resource allocation instruments.
tester	Utilities focused on quality assurance workflows, including test run management and defect submission.
product-owner	Features supporting backlog grooming, roadmap visualization, and feature prioritization.

Environmental Parameter Reference

Parameter	Mandatory	Purpose
TP_DOMAIN	Yes	Host URL fragment (e.g., acme.tpondemand.com)
TP_API_KEY	Conditional*	Preferred authentication token.
TP_USERNAME	Conditional*	Required if using Basic Auth.
TP_PASSWORD	Conditional*	Required if using Basic Auth.
TP_USER_ROLE	Optional	Dictates role-specific tooling (e.g., developer, product-owner).
MCP_STRICT_MODE	Optional	Enforces clean JSON-RPC output for robust client parsing (true).

*One of the authentication methods (Key or Username/Password pair) is mandatory.

Illustrative Conversational Prompts

Examples of directives suitable for the connected AI:

"Retrieve the entire backlog for the 'Phoenix' initiative and summarize any items marked 'Urgent'." "Log 4 hours against story ID 5590, noting completion of the integration testing phase." "Identify all newly reported security defects that lack an assigned owner." "Transition user story #7812 to the 'Ready for QA' state and notify the responsible tester." "Generate a summary report showing bug density trends across the last three sprints."

Comprehensive Documentation Index

• Initial Setup Guide
• Core Operation Concepts
• Tool Signatures Reference
• Advanced Use Cases
• Integration Patterns (LLMs)
• System Architecture Overview