How does the MCP memory integration work?

If enabled, it configures the @modelcontextprotocol/server-memory to store user preferences and project-specific facts that persist globally across all your projects.

Does it support agents other than Claude?

Yes, the infrastructure generated is agent-agnostic and includes configuration support for tools like Cursor, Aider, and Gemini CLI.

Is this skill only for new projects?

No, it is designed to be used on both brand-new repositories and established codebases to improve AI agent efficiency and documentation standards.

Does this skill overwrite my existing project files?

No, the skill performs a silent scan first and explicitly asks for user permission before creating or modifying any files or directories.

What is the purpose of the .agents directory?

The .agents directory stores session state, task progress, and internal agent rules to help AI agents maintain high-quality context across different coding sessions.

Project Initializer for AI Agents

Name: Project Initializer for AI Agents
Author: azusachino

byazusachino

0•

Productivity & Workflow

Bootstraps projects with standardized agent infrastructure, persistent memory protocols, and comprehensive documentation structures.

The init-project skill is a specialized utility designed to prepare any codebase for seamless interaction with AI coding agents. It performs a multi-phase scan to identify tech stacks, tooling gaps, and existing documentation, then generates a robust infrastructure including the 'AGENTS.md' primary reference and a dedicated '.agents' directory. By implementing a standardized Session Management Protocol and integrating with the Model Context Protocol (MCP) for persistent memory, it ensures that Claude and other agents maintain context, track decisions, and follow project-specific conventions across multiple sessions.

Key Features

01Agent-agnostic infrastructure generation including AGENTS.md

02Persistent Session Management Protocol for stateful AI development

030 GitHub stars

04Automated templates for architecture and functional requirements documentation

05Automated project profile scanning (language, framework, build system)

06Model Context Protocol (MCP) server-memory integration for global facts

Use Cases

01Standardizing documentation and task tracking for multi-agent workflows

02Onboarding an AI coding agent to a legacy or newly created repository

03Enabling persistent cross-session memory for complex, long-term development tasks

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add azusachino/harus-skills init-project

For use in Claude.ai and ChatGPT

Download Skill