task-orchestrator and mcp-tasks
About task-orchestrator
jpicklyk/task-orchestrator
A light touch MCP task orchestration server for AI agents. Persistent work tracking and context storage across sessions and agents. Defines planning floors through composable notes with optional gating transitions. Coordinates multi-agent execution without prescribing how agents do their work.
Implements server-enforced workflow gates using YAML-defined note schemas that block phase transitions until required documentation is complete, ensuring deterministic progression independent of agent discipline. Built on MCP's stdio transport with persistent SQLite storage and optional per-project isolation, it surfaces phase-specific structured instructions at exactly the right moment while enabling sub-agent delegation through Claude Code integration. The dependency graph uses typed BLOCKS edges with pattern shortcuts (linear chains, fan-out, fan-in) and BFS traversal to enforce ordering across hierarchical work items up to 4 levels deep.
About mcp-tasks
flesler/mcp-tasks
A comprehensive and efficient MCP server for task management with multi-format support (Markdown, JSON, YAML)
Implements a lightweight 5-tool MCP server designed for LLM task management with automatic WIP limits, duplicate prevention, and optional reminders that remain visible to the AI. Supports stdio and HTTP transports with multi-format persistence (Markdown, JSON, YAML), position-based indexing, and comprehensive search with OR logic filtering. Intentionally restricts AI capabilities to add/move tasks only—preventing rewrites or deletions—while providing CLI access for safe parallel task creation.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work