llm-council - Run high-stakes decisions through an AI council
Runs questions, decisions, code, or plans through five independent AI advisors, anonymous peer review, and a chairman-synthesized verdict with confidence.
Tags
Updated: 2026-10-16Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Frame the decision neutrally
- Enrich questions with workspace context
- Spawn five advisors in parallel
- Review advisor responses anonymously
- Synthesize a final verdict
Inputs
- User question
- Workspace context files
- Referenced files or attachments
- Recent council transcripts
Outputs
- Advisor analyses
- Peer review scores
- Synthesized verdict
- Confidence rating
Requirements
- Claude Code
- Agent tool
- Sub-agent support
- Glob and Read tools