Dart

Creates tasks in Dart project management via the Model Context Protocol by translating AI assistant requests into structured API calls. The server accepts task parameters (title, description, status, priority, assignee, due dates) through MCP tool invocations, validates them against Dart's schema, and persists them via authenticated HTTP requests to the Dart backend using a DART_TOKEN environment variable. This enables AI assistants like Claude, Cursor, and Cline to programmatically create tasks without direct API knowledge.

Creates documents in Dart with structured text content via MCP tool invocations, translating AI-generated content into Dart's document schema. The server accepts document parameters (title, text content, optional folder path) and persists them through authenticated API calls. This enables AI assistants to generate and store documentation, meeting notes, or project specifications directly in Dart's document management system without manual copy-paste workflows.

Provides administrative operations for managing Dart workspaces through MCP tools, enabling privileged operations like user management, workspace configuration, and system administration. These admin tools are exposed through the same MCP interface as regular operations but may require elevated permissions or separate authentication. This enables AI assistants to perform administrative tasks when invoked by authorized users.

Retrieves and filters tasks from Dart using MCP tool invocations with optional status and assignee filters, returning task lists formatted for AI consumption. The server queries the Dart backend via authenticated API calls and can optionally generate AI-friendly summaries of task collections using prompt templates. This enables AI assistants to understand project state, identify blockers, and make context-aware decisions about task creation or updates.

Retrieves Dart workspace configuration (user settings, workspace metadata, API limits) via MCP tool invocation, providing AI assistants with context about the environment they're operating in. The server queries the Dart backend's configuration API and returns structured metadata that helps AI assistants understand constraints, available features, and workspace-specific settings. This enables context-aware behavior — for example, respecting custom task statuses or understanding workspace-specific naming conventions.

Implements the Model Context Protocol using standard input/output (stdio) as the transport mechanism, enabling the server to communicate with any MCP-compatible AI assistant (Claude, Cursor, Cline, Windsurf) without platform-specific code. The server uses the @modelcontextprotocol/sdk package to handle MCP message serialization, request routing, and response formatting over stdio. This architecture allows a single server deployment to serve multiple AI assistants simultaneously through different stdio connections.

Defines reusable MCP prompt templates that guide AI assistants through common Dart operations (create task, create document, summarize tasks) with clear parameter specifications and examples. These prompts are registered with the MCP server and exposed to AI assistants, providing structured guidance on how to invoke tools correctly. The prompts include required/optional parameters, example values, and expected outcomes, reducing the cognitive load on AI assistants and improving consistency of operations.

Defines MCP resource templates that allow AI assistants to discover and retrieve specific Dart entities (tasks, documents) by URI pattern. The server registers resource templates with URI schemes (e.g., `dart://task/{id}`) that enable AI assistants to fetch individual resources by ID without needing to list all resources first. This enables efficient, targeted retrieval and supports resource-based workflows where AI assistants reference specific tasks or documents.

Implements authentication to the Dart backend using a DART_TOKEN environment variable that is passed with every API request. The server reads the token at startup, validates it's present, and includes it in HTTP headers for all downstream API calls to Dart services. This enables secure, stateless authentication without requiring OAuth flows or session management, making the server suitable for automated AI assistant integrations.

Provides Docker containerization for the MCP server, enabling isolated deployment with environment variable injection for authentication and configuration. The server can be deployed as a Docker container with DART_TOKEN and other configuration passed via environment variables, making it suitable for cloud deployments, CI/CD pipelines, and multi-tenant environments. Docker deployment abstracts away Node.js runtime dependencies and provides consistent execution across different host systems.

Enables installation and execution of the MCP server via NPX without requiring local Node.js project setup or package installation. Users can run `npx dart-mcp-server` with environment variables to start the server immediately, making it suitable for quick prototyping and integration with AI assistants. NPX handles downloading the package from npm, managing dependencies, and executing the server in a single command.