convez.ai
MCP ServerIntegração & APIsIntermediário
SerhiiMaksymiv

Targetprocess MCP Server

SerhiiMaksymiv/targetprocess-mcp-server

Integre Targetprocess ao Claude para gerenciar projetos via IA

Ver no GitHub
InicialPluginScore · de 0 a 100

Servidor MCP que integra Claude com a plataforma Targetprocess, permitindo consultas, criação e gerenciamento de entidades como user stories, bugs, tasks e features através de ferramentas estruturadas. Funciona como ponte entre agentes LLM e a API do Targetprocess, com suporte a workflows conduzidos por linguagem natural.

Instalação

$npx convez-ai SerhiiMaksymiv/targetprocess-mcp-server

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
9
Forks
10
Acompanhando
1
Saúde
42%
Último commit
há 2 dias
Versão
—
Issues abertas
1
Contribuidores
6

Casos de uso

  • Gerar casos de teste baseados em user stories e bugs do Targetprocess
  • Consultar releases ativas, features e histórias de usuário em progresso
  • Criar bugs, tasks e comentários vinculados a cartões do Targetprocess
  • Automatizar fluxos de trabalho de gestão de projetos com comandos em linguagem natural

Veredito

MCP server bem estruturado para integração com Targetprocess, com cobertura abrangente de operações CRUD e queries. Ideal para equipes que usam Targetprocess e querem automatizar workflows via Claude, embora exija familiaridade com a plataforma e configuração da API.

README

A tradução para o português deste README ainda está na fila.

Targetprocess MCP Server

An MCP (Model Context Protocol) server that enables AI assistants to interact with the Targetprocess project management platform through structured tools and natural language workflows.

This server exposes powerful capabilities for querying, creating, and managing Targetprocess entities such as user stories, bugs, tasks, and features.

It acts as a bridge between LLM agents and the Targetprocess API, providing:

  • Structured tool-based access
  • Semantic, natural-language-driven workflows

Features

  • Create standalone bugs or bugs linked to a user story/bug card

  • Retrieve TP entities: bugs, user stories, features, releases

  • LLM-Friendly Tools

    • Designed for Claude MCP AI agents

Use Cases examples:

  • "Show me currently active release"
  • "write me test cases based on 145322 tp user story"
  • "write detailed test cases based on 145642 bug, format them inside html element, create a test plan and add test cases to it"
  • "add a comment to 145155 card saying 'test'"
  • "write test cases based on 145640 feature"
  • "create a bug based on 145637 user story where Add Tile flyout (for a Static Tile) not show"
  • "search for a card with 'Text Element' title"

Available tools:

Releases

  • get_current_releases — List all current releases (no params needed)
  • get_release_bugs — Get bugs for a release (name, optional results)
  • get_release_features — Get features for a release (name, optional results)
  • get_release_user_stories — Get user stories for a release (name, optional results)
  • get_release_user_stories_with_description — Same as above but includes full descriptions (name, withDescription)
  • get_release_open_bugs — Get only active/open bugs for a release (name, withDescription, optional results)
  • get_release_open_user_stories — Get only active/open user stories for a release (name, withDescription, optional results)

Features

  • get_feature_user_stories — Get all user stories for a feature by its ID (id)
  • get_not_covered_user_stories_in_feature — Get user stories in a feature not yet covered by tests, includes covered field based on "Test Automation" custom field (id)

User Stories

  • get_user_story_bugs — Get all bugs linked to a user story by its ID (id)

Tasks

  • get_in_progress_tasks_and_bugs — Get all Tasks and Bugs currently in "In Progress" state assigned to a given user (userId)
  • create_task — Create a new task linked to a user story (title, userStoryId, optional description)
  • list_my_user_stories — List User Stories assigned to the current user, optionally filtered by state (optional state, optional take, optional skip)
  • list_my_bugs — List Bugs assigned to the current user, optionally filtered by state (optional state, optional take, optional skip)

Cards — Read

  • get_card_current_status — Get EntityState, TeamState, and assigned teams for a card (id, optional resourceType: UserStory | Bug | Feature, default: UserStory)
  • get_bug_content — Fetch full content of a bug by ID (id)
  • get_user_story_content — Fetch full content of a user story by ID (id)
  • get_bug_comments — Get comments on a bug (id, optional results)
  • get_user_story_comments — Get comments on a user story (id, optional results)
  • get_user_story_test_cases — Fetch the linked test plan and all its test cases (with steps) for a user story (resourceId)
  • get_card_relations — Get all relations (Dependency, Blocker, Relation, Link, Duplicate) for a card, with direction and the related card (id)
  • get_relation_types — List the relation types available in this instance (id + name); use to find the relationType name for create_card_relation (no params)
  • search_tp_cards — Search TP cards by keyword or phrase in description (keyword, optional entityType: UserStories | Bugs, default: UserStories)

Cards — Write

  • add_comment — Post a comment to any card (id, comment)
  • add_comment_with_user — Post a comment to any card and mention a specific user (id, comment, user object from get_users)
  • update_bug — Update an existing bug (id, optional title, optional bugContent, optional origin, optional projectId, optional teamId, optional entityStateId)

    Resolve state name → ID via get_bug_workflows before passing entityStateId

  • update_user_story — Update an existing user story (id, optional title, optional description, optional projectId, optional teamId, optional entityStateId)

    Resolve state name → ID via get_user_story_workflows before passing entityStateId

  • create_card_relation — Create a relation between two cards (masterId, slaveId, optional relationType name, default: "Depends on")

    The Slave depends on the Master — the Master must be done first relationType is matched by name against this instance's types; resolve exact names via get_relation_types (it's resolved to an ID before the API call, since TP rejects relation types passed by name)

  • delete_card_relation — Remove a relation by its relation ID (relationId)

    Get the relationId from get_card_relations first — it's the relation's own ID, not a card ID

  • update_user_story_state — Update the sub-state (team assignment entity state) for a user story (id, optional entityStateId, optional teamId, optional teamAssignmentId)
    1. Call get_user_story_content first to find the assigned team and teamAssignmentId
    2. Call get_user_story_workflows to resolve the target state name → entityStateId
  • create_bug — Create a standalone bug (title, bugContent, optional origin, optional projectId, optional teamId, optional entityStateName)

    origin accepted values: Production - Customer, Production - Internal, Pre-Release - Customer, Pre-Release - Internal, Regression - Dev01, Regression - Team Env, Manual QA (default), Developer Raised, Operations entityStateName accepted values: Backlog, In Triage, Ready for Dev, In Dev, Blocked, PR Raised, Ready for Feature PCH, Ready for Feature QA, In Feature QA, Failed Feature QA, Ready for Merge, Ready to Deploy to Dev01, Ready for Dev01 QA, In Dev01 QA, Failed Dev01 QA, Ready to Deploy to prod, Closed

[!NOTE]
projectId and teamId are optional — fall back to TP_PROJECT_ID and TP_TEAM_ID from config

  • create_bug_based_on_card — Create a bug linked to an existing user story or bug card (card object with id+type, title, bugContent, optional origin, optional projectId, optional teamId)

[!NOTE]
projectId and teamId are optional — fall back to TP_PROJECT_ID and TP_TEAM_ID from config

  • create_user_story — Create a new user story (title, optional description, optional featureId, optional releaseId, optional projectId, optional teamId)

[!NOTE]
projectId and teamId are optional — fall back to TP_PROJECT_ID and TP_TEAM_ID from config

  • create_feature — Create a new feature (title, optional description, optional epicId, optional releaseId, optional projectId, optional teamId)

[!NOTE]
projectId and teamId are optional — fall back to TP_PROJECT_ID and TP_TEAM_ID from config

  • create_test_plan — Create a test plan linked to a UserStory, Bug, or Feature (title, resourceId, optional resourceType, optional description/startDate/endDate)

[!NOTE]
requires TP_PROJECT_ID,

Test Case Workflows

  • write_test_cases — Fetch a card (UserStory, Bug, or Feature) by ID and trigger the full test case writing workflow: Claude analyzes the card, generates detailed test cases covering happy path, edge cases, and error scenarios, creates a linked test plan via create_test_plan, then calls add_test_cases_to_test_plan. Each test case description contains Preconditions and Test Type as HTML; steps are passed as a structured array (resourceId, optional resourceType)
  • add_test_cases_to_test_plan — Add pre-generated test cases to an existing test plan. Each test case has a name, an HTML description (Preconditions and Test Type only), and a steps array of { description, result } objects — steps are created via the TP test step API rather than embedded in the description (testPlanId, testCases array of {name, description, steps})

Processes & Workflows

  • get_processes — Get all Targetprocess processes (no params needed)
  • get_process_workflows — Get workflows for a specific process (processId)
  • get_bug_workflows — Get all bug entity states/workflows for the configured process (no params needed)
  • get_user_story_workflows — Get all user story entity states/workflows for the configured process (no params needed)

Projects

  • get_projects — Get all Targetprocess projects (no params needed)

Teams

  • get_teams — Get all Targetprocess teams returning id and name (no params needed)
  • get_teams_and_team_assignments — Get all teams and team assignments (id and name for each) (no params needed)

User

  • get_logged_in_user — Get the currently logged-in user's info (no params needed)
  • get_users — Get all Targetprocess users (no params needed)
  • get_user_by_id — Get a single Targetprocess user by their ID (id)

Time Tracking

  • log_time — Log time spent on a Task, User Story, or Bug (entityId, entityType: Task | UserStory | Bug, hours, optional description, optional date)
  • get_my_time_logs — Get recent time log entries submitted by the current user (optional take)

Developer Tools

  • get_commit_message — Returns a formatted commit message string for a task or bug ID (id, type: task | bug)

    Format for task on a user story: F#<featureId> US#<userStoryId> T#<taskId> <title> Format for bug on a user story: F#<featureId> US#<userStoryId> B#<bugId> <title> Format for standalone bug: B#<bugId> <title>


Installation

Local Installation for Development

{
  "mcpServers": {
    "targetprocess": {
      "command": "/Users/xxx/.config/nvm/versions/node/v22.14.0/bin/node",
      "args": [
        "/path/to/repository"
      ],
      "env": {
        "TP_TOKEN": "<your-tp-token>" // Settings -> Authentication and Security -> New Access Token,
        "TP_BASE_URL": "<tp-api-endpoint>",
        "TP_OWNER_ID": "<tp-owner-id>", // your user id
        "TP_PROJECT_ID": "<tp-project-id>"
        "TP_TEAM_ID": "<tp-team-id>"
      }
    }
}

Claude Desktop

[!NOTE] You need to have node and npm installed on your machine.

{
  "mcpServers": {
    "targetprocess": {
      "command": "npx",
      "args": [
        "-y",
        "targetprocess-mcp-server"
      ],
      "env": {
        "TP_TOKEN": "<your-tp-token>" // Settings -> Authentication and Security -> New Access Token,
        "TP_BASE_URL": "<tp-api-endpoint>", // required
        "TP_OWNER_ID": "<tp-owner-id>", // required, your user id
        "TP_PROJECT_ID": "<tp-project-id>", // optional see available tools 
        "TP_TEAM_ID": "<tp-team-id>" // optional see available tools
      }
    }
  },
}

Claude Code

claude mcp add targetprocess -s user \
  -e TP_TOKEN=<your-tp-token> -e TP_BASE_URL=<tp-api-endpoint> -- npx -y targetprocess-mcp-server

Local Development

git clone --recursive https://github.com/SerhiiMaksymiv/targetprocess-mcp-server.git
cd targetprocess-mcp-server

npm install
npm run build

Testing

Tests live in tests/ and use Vitest. All tool handlers are extracted to src/handlers/ and tested with mocked TpClient instances — no network calls are made.

npx vitest run        # run all tests once
npx vitest            # watch mode

Coverage

33 of 46 tools (72%) are covered by unit tests.

Test fileHandlers covered
get_bug_content.test.tsget_bug_content
get_user_story_content.test.tsget_user_story_content
get_commit_message.test.tsget_commit_message
get_current_releases.test.tsget_current_releases
get_projects.test.tsget_projects
get_logged_in_user.test.tsget_logged_in_user
get_user_by_id.test.tsget_user_by_id
release_tools.test.tsget_release_user_stories, get_release_bugs, get_release_features, get_release_open_bugs, get_release_open_user_stories
user_team_tools.test.tsget_users, get_teams, get_teams_and_team_assignments
comment_tools.test.tsadd_comment, get_user_story_comments, get_bug_comments
creation_tools.test.tscreate_bug, create_user_story, create_feature, create_task, update_bug, update_user_story_state
my_work_tools.test.tsget_in_progress_tasks_and_bugs, list_my_user_stories, list_my_bugs, log_time, get_my_time_logs
entity_tools.test.tsget_feature_user_stories, get_user_story_bugs, get_card_current_status
Ver o original no GitHub

Bom para

Relacionados