Give Claude Code a one-liner, it builds an expert council and autonomously drives your project to production quality. 10-phase pipeline with adversarial scoring until 100/100.
Agent Claude Code needs its repo
Council Pilot — Research specialist. Discovers expert candidates via web search, collects source URLs, evaluates source tier quality, and produces candidate dossiers ready for distillation.
Agent Claude Code
Council Pilot — Council moderator. Orchestrates expert council deliberation, collects votes, resolves score disputes, and synthesizes consensus scores with gap recommendations.
Agent Claude Code
Council Pilot — Coverage gap analyst. Analyzes scoring reports and expert coverage matrices to identify missing expertise, knowledge gaps, and recommendations for new expert candidates or build focus areas.
Agent Claude Code
Council Pilot — Scoring agent. Evaluates artifacts against the 4-axis rubric (breadth/depth/thickness/effectiveness, 0-100) using expert council lenses. Produces weighted scores with evidence and gap analysis.
Agent Claude Code needs its repo
Council Pilot — Profile filling specialist. Reads source dossiers and fills expert profile contracts using LLM-driven distillation. Produces complete expert profiles with reasoning kernels, critique styles, and knowledge bases.
Agent Claude Code
Council Pilot — Code generation agent. Builds project code guided by expert forum lenses, targeting the weakest scoring axes identified in the latest maturity report. Follows expert reasoning kernels and avoids known anti-patterns.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: