analyze-project

The analyze-project skill is an automated diagnostic tool designed for software engineers and teams working within the Claude OS ecosystem. It performs a deep structural analysis of any project's source code, documentation, and configuration files to build a comprehensive 'project profile.' By standardizing how an AI assistant perceives a codebase, it drastically reduces onboarding time for new developers and ensures that subsequent AI-assisted coding tasks remain consistent with the project's established design patterns and technical requirements.

• Automatically scans local project directories for source files, specifications, and documentation patterns.

• Generates three essential documents: CODING_STANDARDS.md for style and naming conventions, ARCHITECTURE.md for design patterns and module organization, and DEVELOPMENT_PRACTICES.md for workflows and testing strategies.

• Integrates directly with the Claude OS project_profile MCP, allowing the AI to load only necessary context and save memory tokens.

• Supports a wide array of technology stacks, including Rails, Python (Django, FastAPI, Flask), Node.js (NestJS, Next.js), Java (Spring Boot), and Go.

• Enables seamless synchronization with Claude's native memory by creating concise project summaries for easy retention.

• Users can trigger the analysis using natural language commands such as 'analyze-project: [id]' or 'analyze: [path]'.

• Outputs are automatically saved to the .claude-os/project-profile/ directory to maintain a local, version-controlled reference of the project's state.

• For optimal results, execute this skill immediately after initializing a new project or following major refactoring to ensure the AI's documentation stays current with the actual implementation.

• The tool ignores ephemeral directories like node_modules and build artifacts to maintain analysis speed and signal-to-noise ratio.

• Designed for developers who need to keep AI context precise without manual document updates.