agentic-tools-mcp

Agentic Tools MCP Server

Project planning and persistent agent memory come together in agentic-tools-mcp for AI assistants working on development projects. It supports nested tasks, requirements-driven planning, progress inference from code, and searchable project knowledge.

Server facts

agentic-tools-mcp

Project

Author
Pimzino
License
MIT
Category
Productivity

Tech

Runtime / language
node
Transports
stdio
Remote available
No
Authentication
Not specified

Activity

GitHub stars
88
Last commit
Data as of

Overview

Development work in agentic-tools-mcp is organized into projects and nested task trees. Assistants can create and revise tasks, move them between parents, and track dependencies, priority, complexity, status, tags, and estimated or actual hours. Requirements documents can become structured task breakdowns through parse_prd, while recommendation and complexity tools help select upcoming work or identify tasks to split. The infer_task_progress tool examines implementation evidence in the codebase to assess apparent completion rather than relying only on manually updated status. Legacy subtask operations remain available alongside migration to the unified task model.

The memory workflow lets assistants preserve preferences, technical decisions, and project context as categorized records with metadata. They can search those records by text relevance, retrieve details, revise entries, or remove outdated material with confirmation. Research tools guide task-related investigation, generate search queries, and support storing findings as memories. Task and memory files are isolated by working directory by default and can accompany code in Git; an optional global storage mode provides a shared workspace. MCP clients connect through STDIO, and a companion VS Code extension offers a visual interface.

agentic-tools-mcp setup notes

  • agentic-tools-mcp records stdio only; confirm a network transport upstream before planning remote access.

Related MCP servers