council

Coordinate multi-agent debates across personas and rounds into a synthesized decision report.

11|Updated Feb 24, 2026

Search Tags:#decision #multi-agent #orchestration #synthesis #council #debate

Ready to run

npx skills add https://github.com/meanllbrl/dreamcontext --skill council-meanllbrl

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: council
Source: https://github.com/meanllbrl/dreamcontext/tree/main/skill-packs/council
Command: npx skills add https://github.com/meanllbrl/dreamcontext --skill council-meanllbrl

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrates structured multi-agent debates to surface diverse perspectives and produce a final, synthesized decision report for complex questions.

Core Features & Use Cases

Orchestrates 3–10 persona sub-agents across multiple rounds
Produces executive summaries to keep the user context lean
Suitable for product strategy, architecture decisions, vendor selection, post-mortems

Quick Start

Trigger a council debate on a topic to generate a final synthesized decision report.

Frequently Asked Questions about council

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I orchestrate a multi-agent debate for complex product decisions?▼

To orchestrate a multi-agent debate, you trigger a council with a defined topic and 3–10 role personas. The agents debate across multiple rounds, and a synthesizer produces a final executive decision report.

What is structured multi-agent debate and when do I need it?▼

Structured multi-agent debate coordinates 3–10 persona sub-agents to surface diverse perspectives. You need it for complex decisions like product strategy, architecture decisions, vendor selection, or post-mortems requiring well-considered outcomes.

Can I use multi-agent orchestration for vendor selection with multiple personas?▼

Yes, multi-agent orchestration supports vendor selection by assigning 3–10 role personas to debate across multiple rounds. The synthesizer processes these perspectives to produce a final executive summary for your decision.

How many agents can I assign to a multi-agent debate session?▼

You can assign 3 to 10 persona sub-agents to a multi-agent debate session. These agents participate across multiple rounds of discussion before a synthesizer generates a final consolidated decision report.

Does multi-agent debate keep context lean during long orchestration rounds?▼

Multi-agent debate keeps context lean by producing executive summaries instead of relying on raw outputs. The synthesizer ensures only the consolidated decision report is delivered after the multiple debate rounds conclude.

What do I need to start a multi-agent debate for architecture decisions?▼

To start a multi-agent debate for architecture decisions, you need a defined debate topic, role personas, and safeguards ensuring summaries are used. The orchestration handles the rounds and final synthesis automatically.