project-md-refresh

munlucky's avatarfrom munlucky

Analyze the current repository and create or refresh .claude/PROJECT.md to match the project. Use when PROJECT.md is missing, stale, or the user asks to generate/update project-specific rules, structure, commands, or architecture details.

⭐0stars🔀0forks📁View on GitHub🕐Updated Jan 10, 2026
[Developer Tools]

When & Why to Use This Skill

The project-md-refresh skill automates the creation and maintenance of the `.claude/PROJECT.md` file by performing a comprehensive analysis of your repository. It identifies the technology stack, maps directory structures, and extracts essential development commands from configuration files. By ensuring your AI assistant has an accurate, evidence-based 'source of truth,' this skill minimizes context-related errors and optimizes the AI's ability to assist with project-specific coding tasks.

Use Cases

  • Onboarding AI to New Repositories: Automatically generate a detailed project overview, including tech stack and entry points, to give Claude immediate and accurate context.
  • Synchronizing Stale Documentation: Refresh the PROJECT.md file after major refactors or dependency changes to ensure the AI's understanding of the architecture remains up-to-date.
  • Standardizing Development Workflows: Extract and document build, test, and lint commands from files like package.json or Makefile to provide a unified command reference for the agent.
  • Mapping API and Data Patterns: Automatically locate controllers, routes, and models to help the AI understand the data flow and backend structure of the application.
nameproject-md-refresh
descriptionAnalyze the current repository and create or refresh .claude/PROJECT.md to match the project. Use when PROJECT.md is missing, stale, or the user asks to generate/update project-specific rules, structure, commands, or architecture details.

PROJECT.md Refresh

Goal

Create or update .claude/PROJECT.md with accurate, evidence-based project details.

Workflow

  1. Locate the base file.

    • If .claude/PROJECT.md exists, use it as the base; preserve custom rules and update facts.
    • If missing, create .claude/ if needed and copy assets/PROJECT.template.md to .claude/PROJECT.md.

  2. Collect signals from the repository.

    • Read README.md and any project docs.
    • Identify the stack from config files: package.json, pyproject.toml, go.mod, Cargo.toml, build.gradle, pom.xml, Gemfile, requirements.txt, Makefile, or Taskfile.
    • Extract run/build/test/lint commands from scripts or tooling.
    • Inspect the top-level directory structure and key entrypoints.
    • Locate API routes/controllers and data models (search for route, router, controller, handler, model, schema).
    • Find auth configuration (search for auth, jwt, session, oauth).
    • Note environment variable usage (search for ENV, process.env, os.environ, dotenv).

  3. Update PROJECT.md sections.

    • Fill the overview (name, stack, primary language).
    • Summarize core rules and conventions.
    • Document directory structure (top-level + key subdirectories).
    • Document API/data patterns, auth, and docs paths.
    • Add concrete commands for dev/build/lint/test/typecheck.

  4. Output.

    • Save .claude/PROJECT.md.
    • Provide a short summary and list any gaps/questions.

Guardrails

  • Do not invent details; base every statement on files found.
  • If information is missing, add TODOs or ask for confirmation.
  • Keep content concise and project-specific.