Coordinate adversarial brainstorming sessions to refine architectural proposals.
npx skills add https://github.com/kucherenko/gangsta --skill the-grilling
Please help me install this Agent Skill. Skill: the-grilling Source: https://github.com/kucherenko/gangsta/tree/main/skills/the-grilling Command: npx skills add https://github.com/kucherenko/gangsta --skill the-grilling
Enables rigorous architectural critique and refinement through a structured adversarial debate among specialized agents (Proposer, Devils-Advocate, Synthesizer) with Don oversight, ensuring robust designs before Sit-Down.
Initiate a The Grilling session after Reconnaissance completes; the system will coordinate Proposer, Devils-Advocate, and Synthesizer to produce a revised proposal.
High-intent search queries and answers about installing and using this skill.
Adversarial brainstorming is a structured multi-agent debate where a Proposer, Devils-Advocate, and Synthesizer stress-test architectural proposals to surface trade-offs, gather objections, and synthesize refined solutions.
Initiate a session after reconnaissance completes to coordinate the Proposer, Devils-Advocate, and Synthesizer agents. The protocol enforces hard rounds with per-round Don oversight to produce a revised architectural proposal.
You need multi-agent adversarial critique after reconnaissance and before the final Sit-Down to rigorously evaluate architectural risks, resolve alternate solutions, and ensure robust designs through structured debate.
Yes, Autonomous Mode delegates per-round participation to a don-proxy while preserving final authority. It supports autonomous workflows alongside human-in-the-loop oversight within the gangsta framework.
The debate terminates when reaching a Nash Equilibrium, achieving Don consensus, or hitting a predefined round-limit, preventing endless arguments and ensuring the Synthesizer produces a final revised proposal.
yamcodesGroom GitHub issues by eliciting acceptance criteria through interactive questioning.
supertyrelleCompare architectural options and synthesize a recommendation with trade-offs.
kucherenkoAutomate the Heist pipeline from Reconnaissance to Laundering with per-phase don-proxy rules.
AndurilCodeModel players, strategies, payoffs, and timing to diagnose Nash versus Pareto equilibria.
