skills.rest
Kaleb-Rupe
Kaleb-RupeCOMMUNITY
Explore all skills by this author →
𝕏@MightieMagsGitHub

Council

Runs multi-agent debates with visible transcripts where specialized agents respond to each other.

Updated Jul 10, 2026
Search Tags:#architecture#code review#decision making#multi-agent#debate#sanity check#criteria review
One-click install
npx skills add https://github.com/Kaleb-Rupe/aurora --skill council-kaleb-rupe
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Council
Source: https://github.com/Kaleb-Rupe/aurora/tree/main/claude/skills/Thinking/Council
Command: npx skills add https://github.com/Kaleb-Rupe/aurora --skill council-kaleb-rupe

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Single-perspective AI answers often miss blind spots, weak assumptions, and unexplored trade-offs when making architectural or design decisions. This Skill convenes multiple specialized agents (architect, designer, engineer, researcher) that debate a topic in structured rounds, responding to each other's actual points to surface insights through intellectual friction. ## Core Features & Use Cases - Four workflow tiers: Micro (5-10s inline sanity check), Quick (10-20s fast perspectives), ISCReview (15-25s criteria quality review), and Debate (30-90s full 3-round structured debate with transcript). - Visible debate transcripts: Agents reference and challenge each other's specific points across rounds, ending with a synthesis of convergence, disagreements, and a recommended path. - Customizable council composition: Add domain experts like security (Pentester) or a writer when topics touch auth, data, APIs, or public-facing work. - Use Case: Before choosing between WebSockets and SSE for a feature, run the Debate workflow to get a 3-round transcript where the architect, designer, engineer, and researcher challenge each other, then receive a synthesized recommendation. ## Quick Start Ask the AI to run a council debate on whether to use WebSockets or SSE for your real-time feature.

Frequently Asked Questions about Council

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

FAQPage Schema
How do I run a multi-agent debate on a technical decision?▼

Invoke the Debate workflow with your topic, such as "Council: Should we use WebSockets or SSE?" Four agents (architect, designer, engineer, researcher) debate across three rounds in parallel batches, producing a full transcript and synthesis in 30-90 seconds.

What is the difference between Council and RedTeam skills?▼

Council is collaborative-adversarial: agents debate toward the best path with visible multi-round transcripts. RedTeam is purely adversarial: it produces a steelman and counter-argument to attack an idea. Use Council for decisions, RedTeam for stress-testing.

Which Council workflow should I use for a quick sanity check?▼

Use the Micro workflow for inline sanity checks. It runs all four agents in one parallel batch with one-sentence responses, finishing in 5-10 seconds with a Proceed/Pause/Rethink signal. Escalate to Quick or Debate if 2+ agents raise concerns.

Can I add a security expert to the council debate?▼

Yes. Requesting "Council with security" adds the Pentester agent to the debate, which is recommended when topics touch authentication, data handling, or APIs. You can also add a writer agent or restrict the council to specific members.

When should I use ISCReview instead of a full debate?▼

Use ISCReview when evaluating criteria quality rather than choosing between options. Each agent reviews your criteria list for missing items, weak wording, and blind spots in a single 15-25 second round, ideal before the planning phase.

Related Skills

Explore More Software Engineering Skills→
lev-oslev-os

interview

Guide users through structured decision-making with SCAMPER and First Principles frameworks.

Official
Intermediate
NagasakaHNagasakaH

review-plan

Assess dev-plan completeness and coherence against setup.yaml acceptance criteria, outputting review artifacts to docs/{target_repo}/review-plan/.

Community
Advanced
flonatflonat
127

devils-advocate

Challenge research assumptions and identify argument weaknesses through adversarial critique.

Community
Advanced
ruvnetruvnet
70.1k

cognitive-pattern

Define and execute cognitive patterns for agent reasoning and decision-making workflows.

Community
Intermediate