mirror of
https://github.com/K-Dense-AI/claude-scientific-skills.git
synced 2026-03-28 07:33:45 +08:00
6e29728849
Three research-backed cognitive science skills: - consciousness-council: 12-archetype Mind Council deliberation for structured multi-perspective analysis - dhdna-profiler: cognitive fingerprinting across 12 dimensions (DOI: 10.5281/zenodo.18736629) - what-if-oracle: multi-branch scenario analysis with probability weighting (DOI: 10.5281/zenodo.18736841) Each includes reference documentation. All MIT licensed. Source: https://github.com/ashrafkahoush-ux/Claude-consciousness-skills
97 lines
5.1 KiB
Markdown
97 lines
5.1 KiB
Markdown
# Advanced Council Configurations
|
||
|
||
Reference guide for specialized Council configurations beyond the defaults.
|
||
|
||
## Domain-Specific Councils
|
||
|
||
### Startup Decisions
|
||
**Members:** Strategist, Pragmatist, Contrarian, Futurist, Empiricist
|
||
**Why this mix:** Startups need vision (Futurist) grounded in reality (Pragmatist), challenged by skepticism (Contrarian), backed by data (Empiricist), with competitive awareness (Strategist).
|
||
**Key tension to watch:** Futurist vs. Pragmatist — ambition vs. execution capacity.
|
||
|
||
### Technical Architecture
|
||
**Members:** Architect, Minimalist, Empiricist, Outsider, Pragmatist
|
||
**Why this mix:** Architecture needs structure (Architect) that's not over-engineered (Minimalist), validated by evidence (Empiricist), challenged by fresh eyes (Outsider), and actually buildable (Pragmatist).
|
||
**Key tension to watch:** Architect vs. Minimalist — elegance vs. simplicity.
|
||
|
||
### Hiring / People Decisions
|
||
**Members:** Empath, Strategist, Pragmatist, Ethicist, Historian
|
||
**Why this mix:** People decisions need emotional intelligence (Empath), strategic fit (Strategist), practical constraints (Pragmatist), fairness (Ethicist), and pattern recognition (Historian).
|
||
**Key tension to watch:** Empath vs. Strategist — caring for the person vs. optimizing for the team.
|
||
|
||
### Creative Direction
|
||
**Members:** Creator, Outsider, Historian, Empiricist, Minimalist
|
||
**Why this mix:** Creativity needs divergent thinking (Creator), fresh perspective (Outsider), awareness of what's been done (Historian), audience validation (Empiricist), and restraint (Minimalist).
|
||
**Key tension to watch:** Creator vs. Historian — novelty vs. proven patterns.
|
||
|
||
### Crisis Management
|
||
**Members:** Pragmatist, Strategist, Empath, Contrarian, Architect
|
||
**Why this mix:** Crisis needs immediate action (Pragmatist), long-term thinking (Strategist), human awareness (Empath), challenge to groupthink (Contrarian), and systemic fix (Architect).
|
||
**Key tension to watch:** Pragmatist vs. Architect — quick fix vs. root cause.
|
||
|
||
### Ethical Dilemmas
|
||
**Members:** Ethicist, Contrarian, Empath, Historian, Futurist, Empiricist
|
||
**Why this mix (6 members):** Ethical questions deserve more voices. Values framework (Ethicist), challenge to moral certainty (Contrarian), human impact (Empath), precedent (Historian), long-term consequences (Futurist), and evidence (Empiricist).
|
||
**Key tension to watch:** Ethicist vs. Pragmatist (if added) — doing right vs. doing what's possible.
|
||
|
||
### Investment / Financial Decisions
|
||
**Members:** Empiricist, Strategist, Contrarian, Futurist, Pragmatist
|
||
**Why this mix:** Money decisions need data (Empiricist), game theory (Strategist), skepticism of hype (Contrarian), trend awareness (Futurist), and execution reality (Pragmatist).
|
||
**Key tension to watch:** Futurist vs. Empiricist — future potential vs. present evidence.
|
||
|
||
## Custom Archetype Creation
|
||
|
||
Users can define custom archetypes for domain-specific councils. When a user defines a custom member, capture:
|
||
|
||
1. **Name:** What this archetype is called
|
||
2. **Lens:** The primary frame through which they see everything
|
||
3. **Signature question:** The one question they always ask
|
||
4. **Blind spot:** What they consistently miss
|
||
5. **Disagrees with:** Which other archetype they most often clash with
|
||
|
||
**Example custom archetype:**
|
||
```
|
||
Name: The Regulator
|
||
Lens: Compliance and risk management
|
||
Signature question: "What could go wrong legally?"
|
||
Blind spot: Can kill innovation with caution
|
||
Disagrees with: Creator, Futurist
|
||
```
|
||
|
||
## Scoring the Deliberation
|
||
|
||
After synthesis, the Council can optionally score the deliberation quality:
|
||
|
||
| Metric | Scale | What It Measures |
|
||
|--------|-------|-----------------|
|
||
| Diversity Score | 1-5 | How different were the perspectives? (1 = everyone agreed, 5 = genuine disagreement) |
|
||
| Tension Quality | 1-5 | How productive was the central disagreement? (1 = trivial, 5 = illuminating) |
|
||
| Blind Spot Discovery | 1-5 | Did the synthesis reveal something no individual member saw? |
|
||
| Actionability | 1-5 | How concrete and useful is the recommended path? |
|
||
| Overall CQS | 1-5 | Council Quality Score — weighted average |
|
||
|
||
**CQS Formula:** (Diversity × 0.25) + (Tension × 0.30) + (Blind Spot × 0.25) + (Actionability × 0.20)
|
||
|
||
A good deliberation scores 3.5+ overall. Below 3.0, consider re-running with different members or a reframed question.
|
||
|
||
## Multi-Round Deliberation
|
||
|
||
For complex questions, enable "Rounds Mode":
|
||
|
||
**Round 1:** Initial positions (standard deliberation)
|
||
**Round 2:** Each member responds to the member they most disagree with
|
||
**Round 3:** Revised positions after hearing counterarguments
|
||
**Final Synthesis:** Incorporates all rounds
|
||
|
||
Multi-round deliberation produces deeper insight but takes longer. Use for high-stakes decisions where the extra depth is worth it.
|
||
|
||
## Silent Council Mode
|
||
|
||
Sometimes the user doesn't need the full deliberation output — they just need the synthesis. In "Silent Council" mode:
|
||
|
||
1. Run the full deliberation internally
|
||
2. Only output the Synthesis section
|
||
3. Offer to "show the full deliberation" if the user wants the reasoning
|
||
|
||
This is faster and less overwhelming for quick decisions.
|