multi-brain-experts
Multi-Brain Experts Protocol
Instead of fixed Creative/Pragmatic/Comprehensive perspectives, dynamically select 3 domain experts from a role pool based on the request context. Each expert evaluates the request through their specialized lens.
Workflow
1. Understand the request
2. Select 3 experts from the role pool
3. Each expert provides their perspective (2-3 sentences)
4. Synthesize consensus
5. Produce full output with all perspectives visible
Step 1: Understand the Request
Identify the domain, constraints, and key decision factors. This determines which experts to activate.
Step 2: Expert Selection
Choose the 3 most relevant experts from the role pool. Selection criteria:
- Relevance: How directly does this role address the core problem?
- Coverage: Do the 3 roles cover different dimensions (technical, user-facing, business)?
- Tension: Prefer combinations that naturally create productive tension.
See references/EXPERT_ROLES.md for the full role pool and selection heuristics.
Selection Output
Always declare the selected experts before their perspectives:
**Selected Experts:** [Role A], [Role B], [Role C]
**Why these 3:** [One sentence explaining the selection logic]
Step 3: Expert Perspectives
Each expert provides their analysis independently:
## 🧠 Expert Panel
**Selected Experts:** Security Architect, Performance Engineer, DX Advocate
**Why these 3:** API design with auth requires security-first thinking, latency awareness, and developer ergonomics.
**🔒 Security Architect:**
[2-3 sentences from security perspective]
**⚡ Performance Engineer:**
[2-3 sentences from performance perspective]
**🛠️ DX Advocate:**
[2-3 sentences from developer experience perspective]
Step 4: Consensus
Synthesize expert opinions:
- Agreement points: Where experts align
- Complementary insights: Unique contributions from each expert
- Conflicts: Which expert's concern takes priority and why
Step 5: Full Output
Mandatory: The final response must always include the selected experts, all perspectives, the consensus, and the complete deliverable.
Guardrails
- Always show expert selection reasoning — the user must understand why these 3 were chosen.
- Each expert must reason within their domain — no generic advice.
- If the request is purely within one domain, still select 3 experts but from adjacent disciplines.
- Prefer productive tension over agreement — complementary expertise is more valuable than consensus-seeking.
- Fall back to base
multi-brain(Creative/Pragmatic/Comprehensive) if no clear domain experts apply.
References
- See
references/EXPERT_ROLES.mdfor the complete role pool with descriptions and trigger conditions. - See
references/EXAMPLES.mdfor worked examples.
More from fatih-developer/fth-skills
task-decomposer
Break down large, complex, or ambiguous tasks into independent subtasks with dependency maps, execution order, and success criteria. Plan first, then execute step by step. Triggers on 'how should I do this', 'where do I start', 'plan the project', 'break it down', 'implement' or whenever a task involves multiple phases.
24context-compressor
Compress long conversation histories, large code files, research results, and documents by 70% without losing critical information. Triggers when context window fills up, when summarizing previous steps in multi-step tasks, before loading large files into context, or on 'summarize', 'compress', 'reduce context', 'save tokens'.
18multi-brain-debate
Two-round debate protocol where perspectives challenge each other before consensus. Round 1 presents independent positions, Round 2 allows counter-arguments and rebuttals. Produces battle-tested decisions for high-stakes choices.
17multi-brain-score
Confidence scoring overlay for multi-brain decisions. Each perspective rates its own confidence (1-10) with justification. Consensus uses scores as weights, flags low-confidence areas, and surfaces uncertainty explicitly.
15checkpoint-guardian
Automatic risk assessment before every critical action in agentic workflows. Detects irreversible operations (file deletion, database writes, deployments, payments), classifies risk level, and requires confirmation before proceeding. Triggers on destructive keywords like deploy, delete, send, publish, update database, process payment.
14parallel-planner
Analyze multi-step tasks to identify which steps can run in parallel, build dependency graphs, detect conflicts (write-write, read-write, resource contention), and produce optimized execution plans. Triggers on 3+ independent steps, 'speed up', 'run simultaneously', 'parallelize', 'optimize' or any task where sequential execution wastes time.
14