Plugins

Framework multi-agente de resolução e verificação de problemas matemáticos: brainstorming → iteração de solucionador → debate multi-verificador → base de conhecimento verificada, com retomada por checkpoint e intervenção manual/automática (também distribuído como preset de agente).

Análise multi-agente de ações A-share atrás de /trading-agent, portada de TradingAgents-AShare, executando 14 papéis (analistas, debate bull/bear, revisão de risco, trader) através da rota LLM do harness.

Runs a thesis-antithesis-seal debate whose verdict is computed from real command exit codes and case counts, never from agent prose.

Council of High Intelligence: monte um think tank de 18 figuras históricas para debate estruturado em múltiplas perspectivas, com um painel lateral DAG ao vivo, histórico de execução em SQLite e padrões por projeto via .council.yaml.

Delega tarefas, revisões e debates a modelos alternativos (DeepSeek, MiniMax, Ollama, etc.) através do CLI gateway-companion, sem MCP.

Answers hard questions by staging a formal debate: two adversarial subagents argue opposite sides with research, then the main agent rules on the transcript through the debate_open, debate_round, and debate_verdict tools.

Multi-role parallel argumentation for DeepSeek Harness: runs the real Codex and Claude CLIs in parallel, has the DSH model (Judge) write a third-party verdict back into the conversation, plus Obsidian-style single-agent direct chat and per-role model config.

Multi-agent isolated debate for DeepSeek Harness: debaters of different temperaments research independently (no herding), then cross-examine, then judges converge to a conclusion. Runs as a background job with per-agent progress, token accounting, wave-parallel execution, and session-referenced resume.