council-consensus
用于Codex和Gemini探索不同观点的多角度分析。在"委员会观点"、"委员会优缺点"、"委员会评估选项"、"权衡是什么"、"委员会共识"时触发。
📖 中文导读:用于Codex和Gemini探索不同观点的多角度分析。在"委员会观点"、"委员会优缺点"、"委员会评估选项"、"权衡是什么"、"委员会共识"时触发。 以下正文为该技能的原始 SKILL.md 内容(英文),上方安装命令可直接使用。
name: council-consensus description: Use for multi-perspective analysis with both Codex and Gemini exploring different viewpoints. Triggers on "council perspectives", "pros and cons from council", "council evaluate options", "what are the tradeoffs", "council consensus".
Council Consensus Skill
Multi-perspective analysis with both Codex and Gemini, each exploring for/against/neutral viewpoints for comprehensive evaluation.
When to Use
- Evaluating technology choices
- Architecture decisions
- Trade-off analysis
- When user needs pros and cons
- Comparing different approaches
- Decision-making support
Reasoning Level
high (default for balanced analysis)
Execution
Identify the topic/question for consensus
Gather relevant context
For each model, request stance-based analysis:
Analyze this topic from multiple perspectives: Topic: <topic or decision> Please provide: 1. **FOR** arguments - Best case for this approach 2. **AGAINST** arguments - Concerns and risks 3. **NEUTRAL** assessment - Objective analysis 4. **Recommendation** - Your balanced conclusionRun BOTH commands in parallel:
Codex:
codex exec --sandbox read-only -c model_reasoning_effort="high" "<prompt>"Gemini:
gemini -s -y -o json "<prompt>"Synthesize multi-model, multi-perspective analysis
Response Format
## AI Council Consensus Analysis
### Codex (GPT-5.2) Perspectives:
**FOR:**
- [Pro arguments]
**AGAINST:**
- [Con arguments]
**NEUTRAL:**
- [Objective observations]
**Recommendation:** [Codex's conclusion]
---
### Gemini Perspectives:
**FOR:**
- [Pro arguments]
**AGAINST:**
- [Con arguments]
**NEUTRAL:**
- [Objective observations]
**Recommendation:** [Gemini's conclusion]
---
### Council Consensus:
**Unanimous FOR:** [Points both models agree are advantages]
**Unanimous AGAINST:** [Concerns both models share]
**Divergent Views:** [Where the models disagree]
**Cross-Model Insights:**
- [Unique insights from combining perspectives]
**Council Recommendation:**
[Synthesized recommendation based on all perspectives from both models]
**Confidence Level:** [High if both agree, Medium if partial agreement, Low if divergent]
---
*Session IDs: Codex=[id], Gemini=[id]*