A community-driven registry for Claude, Cursor, Windsurf, Cline & more. Not affiliated with Anthropic.
Are you the author? Sign in to claim
A local Model Context Protocol (MCP) server providing backend tools for client-driven project and task management using
A local Model Context Protocol (MCP) server providing backend tools for client-driven project and task management using a SQLite database.
This server acts as a persistent backend for local MCP clients (like AI agents or scripts) that need to manage structured task data within distinct projects. It handles data storage and provides a standardized set of tools for interaction, while the strategic workflow logic resides within the client.
Key Features:
./data/taskmanager.db by default) for simple, self-contained data storage.The following tools are available for MCP clients:
createProject:
projectName (string, optional, max 255){ project_id: string }addTask:
project_id (string, required, UUID), description (string, required, 1-1024), dependencies (string[], optional, max 50), priority (enum 'high'|'medium'|'low', optional, default 'medium'), status (enum 'todo'|'in-progress'|'review'|'done', optional, default 'todo')TaskData object of the created task.listTasks:
project_id (string, required, UUID), status (enum 'todo'|'in-progress'|'review'|'done', optional), include_subtasks (boolean, optional, default false)TaskData or StructuredTaskData objects.showTask:
project_id (string, required, UUID), task_id (string, required)FullTaskData object.setTaskStatus:
project_id (string, required, UUID), task_ids (string[], required, 1-100), status (enum 'todo'|'in-progress'|'review'|'done', required){ success: true, updated_count: number }expandTask:
project_id (string, required, UUID), task_id (string, required), subtask_descriptions (string[], required, 1-20, each 1-512), force (boolean, optional, default false)FullTaskData object including new subtasks.getNextTask:
project_id (string, required, UUID)FullTaskData object of the next task, or null if none are ready.exportProject:
project_id (string, required, UUID), format (enum 'json', optional, default 'json')importProject:
project_data (string, required, JSON), new_project_name (string, optional, max 255){ project_id: string } of the newly created project.updateTask:
project_id (string, required, UUID), task_id (string, required, UUID), description (string, optional, 1-1024), priority (enum 'high'|'medium'|'low', optional), dependencies (string[], optional, max 50, replaces existing)FullTaskData object.deleteTask:
project_id (string, required, UUID), task_ids (string[], required, 1-100){ success: true, deleted_count: number }deleteProject:
project_id (string, required, UUID){ success: true }(Note: Refer to the corresponding src/tools/*Params.ts files for detailed Zod schemas and parameter descriptions.)
Prerequisites: Node.js (LTS recommended), npm.
Install Dependencies:
npm install
Run in Development Mode: (Uses ts-node and nodemon for auto-reloading)
npm run dev
The server will connect via stdio. Logs (JSON format) will be printed to stderr. The SQLite database will be created/updated in ./data/taskmanager.db.
Build for Production:
npm run build
Run Production Build:
npm start
DATABASE_PATH environment variable. The default is ./data/taskmanager.db.LOG_LEVEL environment variable (e.g., debug, info, warn, error). The default is info./src: Source code.
/config: Configuration management./db: Database manager and schema (schema.sql)./repositories: Data access layer (SQLite interaction)./services: Core business logic./tools: MCP tool definitions (*Params.ts) and implementation (*Tool.ts)./types: Shared TypeScript interfaces (currently minimal, mostly in repos/services)./utils: Logging, custom errors, etc.createServer.ts: Server instance creation.server.ts: Main application entry point./dist: Compiled JavaScript output./docs: Project documentation (PRD, Feature Specs, RFC)./data: Default location for the SQLite database file (created automatically).tasks.md: Manual task tracking file for development.package.json, tsconfig.json, .eslintrc.json, etc.)npm run lintnpm run format(Code is automatically linted/formatted on commit via Husky/lint-staged).
MCP server integration for DaVinci Resolve Studio
mcp-language-server gives MCP enabled clients access semantic tools like get definition, references, rename, and diagnos
Run Claude Code as an MCP server so any agent can delegate coding tasks to it
Browser automation using accessibility snapshots instead of screenshots