Premortem Runner

Most plans fail in ways someone could have predicted.

Premortem Runner is an OpenClaw skill that facilitates a structured adversarial review before you ship, commit, or go public. Unlike a blank-page "what could go wrong" exercise, it runs three distinct hostile personas — Saboteur, New Customer, and Auditor — and scores every failure mode they surface by likelihood × impact so you know what to fix first.

What it does

• Runs three adversarial perspectives on your project brief: execution risks, market/user risks, and compliance/reputational risks
• Scores each failure mode (likelihood × impact) and ranks the full list so the highest-priority items surface immediately
• Generates concrete mitigations with suggested owners and trigger conditions for the top 5 risks
• Saves a "what was foreseen" artifact alongside your project doc — if a post-mortem runs later, you'll know what you saw coming

What's in the package

• SKILL.md — the full facilitation workflow your agent follows, including persona definitions, scoring rubric, and output format.

How it ships

Drop the folder into ~/.openclaw/skills/premortem-runner/. Restart OpenClaw. Run /premortem-runner or just say "run a premortem on [project]" — your agent picks it up.

Pairs well with Reagan (executive assistant persona) for high-stakes decisions and pre-board prep.

Questions: hackedcubeapps@gmail.com (response within 48 hours).