What problem does it solve? Understanding an unfamiliar project directory normally requires opening files one by one, which wastes time and context. This Skill produces a deterministic inventory report covering the tree, languages, manifests, git hotspots, and entry points so you know what is worth reading before you start. ## Core Features & Use Cases - Deterministic project inventory: Runs scripts/pop_recon.py (stdlib only, zero LLM) to produce a RECON.md with tree, languages/LOC, manifests, git churn hotspots, and entry points/configs/CI. - Writing mode for markdown bases: For mostly-markdown directories, the report switches to chapter/heading structure, wordcounts, and a frontmatter inventory instead of code metrics. - Use Case: At the start of a delegated recon or during task planning that requires broad reading, generate the RECON.md of the target directory, pass it to subagents as context, and only then open the specific files the report could not answer. ## Quick Start Run the pop_recon.py script on the target project directory and read the generated RECON.md report in full before opening any project file.
