Summary

This project provides an intelligent development framework for Claude Code, transforming it into a project-aware AI programming assistant. It addresses the limitations of generic AI chat models by integrating with local LLM engines like Ollama to automatically detect project context, analyze user intent, and recommend specialized AI "Skills." The primary benefit is a streamlined, automated development workflow, enabling users to achieve end-to-end development automation from requirements to delivery with minimal manual configuration.

How It Works

The framework operates via a Python Hooks system that intercepts and automates actions throughout the Claude Code session lifecycle. Core intelligence is derived from locally deployed Ollama, which handles project type detection (e.g., Python/FastAPI), dependency identification, and sophisticated intent analysis. This enables the system to dynamically load over 300 pre-defined "Skills"—specialized AI agents for backend, frontend, testing, security, and DevOps tasks. A key design choice is the use of automatically maintained documentation files (DEVELOPMENT.md, KNOWLEDGE.md, CHANGELOG.md) for project context, bypassing traditional memory limitations and preventing context explosion. Prompt templates are centralized in prompts.json for easy customization and version control.

Quick Start & Requirements

• Prerequisites: Claude Code (Desktop or VS Code extension), Ollama (recommended), uv (recommended).
• Installation: Install Ollama and a model (e.g., gemma3:1b), install uv, clone the repository, configure .env if necessary, and open the project directory in Claude Code.
• Links: Quick start guide is integrated within the README (# - 5分钟快速开始). Detailed documentation is available locally at ./project_document.

Highlighted Details

• 300+ Specialized Skills: Offers expert AI agents for backend, frontend, testing, security, DevOps, and more, automatically discoverable and loadable.
• Automated Documentation: Enforces continuous updates to DEVELOPMENT.md, KNOWLEDGE.md, and CHANGELOG.md, ensuring documentation stays synchronized with code changes.
• Intelligent Intent Analysis: Automatically assesses task complexity and recommends appropriate tools and Skills, reducing manual prompt engineering.
• Zero-Configuration Startup: Designed for immediate use upon cloning; configurations are managed via Hooks.
• Local Intelligence: Leverages Ollama for all decision-making, ensuring privacy and offline capability.

Maintenance & Community

The project is actively developed by @Prorise-cool, with contributions welcomed via GitHub Issues and Discussions. The roadmap indicates planned multi-language support in v1.1.0. No specific community channels like Discord or Slack are listed.

Licensing & Compatibility

The project is licensed under the MIT License, which permits broad use, modification, and distribution, including for commercial purposes and integration into closed-source projects.

Limitations & Caveats

The system's effectiveness relies heavily on the correct setup and performance of Claude Code and Ollama. Project detection accuracy is dependent on Ollama's capabilities. While not explicitly stated as alpha, the project is under active development with a roadmap for future enhancements. The primary author is listed, suggesting a potential bus factor.