What problem does it solve? When entering an unfamiliar codebase, developers waste time manually reading READMEs, package manifests, and directory trees just to understand what a project is and how to run it. This Skill automates that reconnaissance and delivers a single structured project card instead of raw file dumps. ## Core Features & Use Cases - Cost-progressive collection: Gathers information in a 6-step order (ls → package manifest → README → directory skeleton → git log → rg scan), stopping as soon as enough data exists to produce the card. - Structured project card output: Covers identity, tech stack (language, framework, runtime, package manager, deployment target), directory skeleton, key modules, build/run commands, recent activity, and known constraints. - Edge-case handling: Supports monorepos, microservices, CLI tools, missing manifests, lockfile-only projects, and outdated READMEs with explicit "Not found" or "Inferred" annotations instead of guesses. - Use Case: You clone a Go monorepo and ask "what is this project?" — the Skill detects sub-package manifests, runs tiered collection, and outputs a root overview card plus per-sub-package tech stack cards. ## Quick Start Ask the assistant to analyze the current project and produce a project overview card with its tech stack and architecture.
