菜单
Meta-orchestrator that selects optimal reasoning patterns based on problem characteristics. Analyzes 11 dimensions and recommends ToT, BoT, SRC, HE, AR, DR, AT, RTR, or NDF patterns.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Self-improving agent architecture using ChromaDB for continuous learning, self-evaluation, and improvement storage. Agents maintain separate memory collections for learned patterns, performance metrics, and self-assessments without modifying their static .md configuration.
Guide for manually creating custom skills from scratch with templates and validation. Use when designing NEW skills with specific requirements. For auto-generating skills from existing git repos, use /generate-skills instead.
Display QAVR (Q-Value Augmented Vector Retrieval) system status including mode, interaction counts, top memories by Q-value, and configuration.
Enables continuous self-improvement through learning from failures, user corrections, and capability gaps. Integrates with QAVR for learned memory ranking.
Stress-test solutions using the STRIKE framework. Systematically attack proposals to find weaknesses before deployment.
Exhaustive solution space exploration. Use when you need ALL viable options (not just the best), solution space is unknown, or you can't afford to miss alternatives.
| name | integrated-reasoning-v2 |
| description | Meta-orchestrator that selects optimal reasoning patterns based on problem characteristics. Analyzes 11 dimensions and recommends ToT, BoT, SRC, HE, AR, DR, AT, RTR, or NDF patterns. |
| metadata | {"clawdbot":{"requires":{"env":[]}}} |
| user-invocable | true |
| homepage | https://github.com/anthropics/claude-code |
Meta-orchestration guide for choosing optimal reasoning patterns. Analyzes problem characteristics and recommends which cognitive methodology to use.
Invoke when facing complex problems and unsure which reasoning approach fits best. IR-v2 scores 11 dimensions and selects the optimal pattern.
| Pattern | Best For | Key Strength |
|---|---|---|
| ToT (Tree of Thoughts) | Find THE best among known options | Systematic evaluation |
| BoT (Breadth of Thought) | Explore ALL possible approaches | Exhaustive coverage |
| SRC (Self-Reflecting Chain) | Step-by-step with validation | Error correction |
| HE (Hypothesis-Elimination) | Find root CAUSE of issues | Diagnostic precision |
| AR (Adversarial Reasoning) | Stress-test solutions | Robustness validation |
| DR (Dialectical Reasoning) | Synthesize opposing views | Trade-off resolution |
| AT (Analogical Transfer) | Novel problems, no precedent | Creative solutions |
| RTR (Rapid Triage Reasoning) | Time-critical decisions | Speed optimization |
| NDF (Negotiated Decision Framework) | Multi-stakeholder alignment | Consensus building |
"I need the BEST option" → ToT
"I need ALL options" → BoT
"I need to trace step by step" → SRC
"I need to find THE CAUSE" → HE
"I need to VALIDATE a solution" → AR
"I'm stuck between two approaches" → DR
"This is completely novel" → AT
"I need to decide NOW" → RTR
"Multiple stakeholders must agree" → NDF
Before selecting a pattern, score these dimensions:
ToT = (CriteriaClarity × 0.35) + (SingleAnswer × 0.30) + (SolutionSpaceKnown × 0.20) + (SequentialDeps × 0.15)
BoT = (1-SingleAnswer × 0.35) + (ProblemNovelty × 0.30) + (1-CriteriaClarity × 0.20) + (SequentialDeps × 0.15)
SRC = (SequentialDeps × 0.45) + (CriteriaClarity × 0.25) + (EvidenceAvailable × 0.20) + (RobustnessRequired × 0.10)
HE = (EvidenceAvailable × 0.40) + (CriteriaClarity × 0.30) + (SequentialDeps × 0.20) + (RobustnessRequired × 0.10)
AR = SolutionExists >= 3 ? (RobustnessRequired × 0.45) + (EvidenceAvailable × 0.25) + (CriteriaClarity × 0.20) + (1-TimePressure × 0.10) : 0
DR = (OpposingValidViews × 0.50) + (CriteriaClarity × 0.20) + (1-SingleAnswer × 0.15) + (1-TimePressure × 0.15)
AT = (ProblemNovelty × 0.45) + (1-SolutionSpaceKnown × 0.30) + (1-CriteriaClarity × 0.15) + (EvidenceAvailable × 0.10)
RTR = (TimePressure × 0.50) + (CriteriaClarity × 0.25) + (RobustnessRequired × 0.15) + (SequentialDeps × 0.10)
NDF = StakeholderComplexity >= 3 ? (StakeholderComplexity × 0.50) + (OpposingValidViews × 0.25) + (1-TimePressure × 0.15) + (CriteriaClarity × 0.10) : 0
When top 2 patterns score within 0.3 of each other:
User: "Should we use microservices or monolith for our new app?"
IR-v2 Analysis:
- Sequential Dependencies: 2 (independent evaluation)
- Criteria Clarity: 3 (some clear, some subjective)
- Solution Space Known: 5 (two options)
- Single Answer Needed: 5 (must pick one)
- Opposing Valid Views: 4 (both have merit)
Top Patterns:
1. DR = 3.85 (opposing valid views present)
2. ToT = 3.60 (clear options to evaluate)
Recommendation: Use DR to synthesize trade-offs, then ToT to score final decision.
When invoked, provide: