Get a quick architecture summary of an unfamiliar repo from the terminal
Generate a CODEMAPPR.md file for a pull request or repo docs
Export a dark-themed HTML overview with a collapsible directory tree
Skip cache or temp folders during a scan with --ignore patterns
Single pip install with no API key or network call needed.
CodeMappr is a command-line tool that prints a quick, high-level summary of any codebase it is pointed at. You install it from PyPI with pip install codemappr, then run codemappr scan in a project folder. It walks the directory tree, makes a guess about what kind of project it is, lists the languages used, and produces a short architecture overview. The README's main selling point is that the tool runs entirely on your machine and does not need an API key or an internet connection. Project type detection is rule-based. The README lists over twenty patterns the scanner looks for: a packages or apps folder for monorepos, package.json with React or Next for Next.js.vue files for Vue, angular.json for Angular, manage.py with settings.py and urls.py for Django, FastAPI or Flask entry points, pyproject.toml or setup.py for Python libraries, Cargo.toml for Rust, go.mod for Go, Maven pom.xml or Gradle build files for Java, AndroidManifest.xml for Android, pubspec.yaml for Flutter, Gemfile and routes.rb for Rails, composer.json with artisan for Laravel, Dockerfile or docker-compose.yml.ipynb notebooks with pandas for data science, a root index.html for static sites, and generic Python or JS fallbacks. There are three output modes. The default is an in-terminal dashboard with colored panels for the project profile, the architecture summary, and a styled directory tree. Passing --format md writes a portable CODEMAPPR.md file you can drop into a pull request or repository docs. Passing --format html writes a single CODEMAPPR.html with a dark theme and collapsible directory trees, meant for browsing larger codebases. Passing --format all writes every format in one run. Other CLI options in the quick start include a path argument to scan a specific directory, --depth to cap how far the tree walk goes, and --ignore with a comma-separated list of patterns to skip such as .cache or temp. The README does not describe the heuristics beyond the detection-signal table. The roadmap lists three milestones. Version 1.0.0 is already done and covers the core scanning, detection, terminal output, and exports. Version 2.0.0 is planned to add a file relationship map, and version 3.0.0 is planned to add a structural diff or changelog feature. Contributions are accepted through the usual fork, branch, run pytest, and pull request flow, and new features are expected to come with tests and doc updates. The license is MIT.
Generated 2026-05-22 · Model: sonnet-4-6 · Verify against the repo before relying on details.