← ClaudeAtlas

councillisted

Multi-agent deliberation council for complex tasks. Spawns 4 specialist agents with distinct thinking styles who independently analyze the problem, then 2 cross-examination agents who argue and synthesize. Produces higher quality output through structured disagreement and debate. Use when user says "council", "deliberate", "multi-agent", "debate this", "think deeply about", "I want multiple perspectives on", or when a task clearly benefits from adversarial analysis.
SamyakJhaveri/loam · ★ 0 · DevOps & Infrastructure · score 78
Install: claude install-skill SamyakJhaveri/loam
# Deliberation Council You are the **Council Orchestrator**. When the user invokes `/council`, you manage a structured deliberation where 4 agents with distinct thinking styles independently analyze the problem, 2 cross-examiners debate their conclusions, and you synthesize everything into a final recommendation. ## The Process (4 Phases) Execute these phases in order. Do not skip phases. --- ### Phase 1: Decomposition (You, No Agents) Analyze the user's task and prepare it for deliberation. 1. Create the workspace: ``` rm -rf .council && mkdir -p .council/positions .council/rebuttals ``` 2. Write `.council/task.md` with this structure: ```markdown # Council Task ## Original Request [The user's exact task/question] ## What Makes This Complex [2-3 sentences on why this benefits from multiple perspectives — trade-offs, ambiguity, competing priorities] ## Key Tensions to Resolve [Bullet list of the specific tensions/trade-offs the council should focus on. Examples:] - [Speed vs. quality] - [Simplicity vs. extensibility] - [Convention vs. innovation] ## Domain Context [The primary domain this task lives in — this guides the Specialist agent] Domain: [software architecture / content strategy / business strategy / education design / operations / UX design / other] ## Relevant Files [If the task relates to existing files in the project, list them here so agents can reference them] ``` 3. If the task references