What problem does it solve? Releasing a private project publicly risks leaking secrets, credentials, PII, and internal references. This Skill automates a three-stage pipeline that forks the project, strips sensitive data, verifies sanitization, and generates release-ready packaging. ## Core Features & Use Cases - Three-Agent Pipeline: Chains a forker (copies files, strips secrets, generates .env.example), a sanitizer (six scan categories with PASS/FAIL verdict), and a packager (CLAUDE.md, setup.sh, README, LICENSE, CONTRIBUTING.md). - Safety Gate with Retry: Blocks publishing on sanitizer FAIL, supports up to three fix-and-rescan attempts, and never pushes to GitHub without explicit user approval. - Staging Management: Tracks staged projects under $HOME/opensource-staging with list and status commands showing FORK_REPORT.md and SANITIZATION_REPORT.md. - Use Case: You have a private internal tool you want to publish. Run the full pipeline to strip API keys, verify no PII remains, generate documentation and a license, then create the public GitHub repo with one approval. ## Quick Start Ask the AI to open source your project by saying "open source this project" and answer the prompts for license, GitHub org, and description.
