Skill Claude CodeCodex
Run a query through a multi-model LLM Council with configurable models. Default: Claude Opus, Gemini 3.1 Pro, GPT-5.4. Each model answers independently, then Claude synthesizes a verdict. With 4+ models, anonymous peer review runs before synthesis. Trigger: /llm-council-crossmodel.
Skill Claude CodeCodex
Use when assessing whether a GitHub PR has sufficient information to reproduce and verify its changes. Checks for linked issues, reproduction steps, expected behavior, test evidence, and deployment notes. Takes a GitHub PR URL as input and produces a readiness report with pass/fail verdicts per criterion.
Skill Claude CodeCodex
Generates manual step-by-step verification instructions for confirming a specific GitHub PR fix is correctly included in a DWO prerelease build. Takes a GitHub PR URL as input and produces a ready-to-run checklist for the user to execute in their terminal.