New community skill: Root Cause Analyst for analyzing why AI coding sessions struggle

Built a deep-dive diagnostic skill that turns your brain/ folder artifacts into actionable insights.

It doesn’t just count revisions — it classifies:

  • Scope changes: human-added vs necessary-discovered vs agent-overreach
  • Rework shapes: clean, early replan, progressive expansion, late verification churn, etc.
  • Root causes: SPEC_AMBIGUITY, REPO_FRAGILITY, AGENT_ARCHITECTURAL_ERROR, etc. (with evidence + confidence)
  • Friction hotspots by subsystem/file
  • Non-obvious findings
  • Auto-generates prompt_improvement_tips.md + updates project-health-state

PR is open here: Add analyze-project — Root Cause Analyst skill by Gizzant · Pull Request #297 · sickn33/antigravity-awesome-skills · GitHub

Full SKILL.md + workflow inside. Test it locally by dropping into ~/.gemini/antigravity/.agent/skills/analyze-project/

Feedback / PRs welcome — especially if you’ve got sessions that keep derailing and want to see why!

Hello,

Thank you for this insightful feedback! We’re thrilled to hear that Antigravity is helping you build more efficiently and professionally. We appreciate you being part of the community and are excited to see your progress!