What problem does it solve? Solo founders and indie hackers lack teammates to review plans, test code, audit design, and manage releases. This Skill gives an AI coding agent a set of structured specialist roles — CEO reviewer, engineer, designer, QA lead, and release engineer — so one person can run a full product development loop from idea validation to deployment. ## Core Features & Use Cases - Full development loop: Skills for /validate, /plan, /review, /qa, /ship, /document-release, and /retro cover the entire build-ship-reflect cycle. - Persistent headless browser: A compiled Bun CLI drives a long-lived Chromium daemon over localhost HTTP, giving sub-second browser commands with ref-based element selection for QA and design review. - Template-generated skill docs: SKILL.md files are generated from .tmpl templates with placeholders resolved from source code, keeping agent instructions in sync with actual commands. - Use Case: A solo founder runs /validate to test demand for an idea, /plan to review architecture and design, /qa to click through the app in a real browser and fix bugs, then /ship to test, review, and open a PR — all in one terminal session. ## Quick Start Ask the agent to run the /validate skill to stress-test your product idea before writing any code.