# Reasoning Protocols
**From Agentic Orchestration to Reasoning Frameworks**

## Overview

This directory contains **Reasoning Protocols** — structured frameworks for guiding thought rather than orchestrating agents. These protocols transform the Worksona Delegator's agentic patterns into cognitive frameworks that guide reasoning processes.

### The Transformation

**Before**: Agentic Orchestration
- Delegator creates specialized agents
- Agents execute tasks independently
- Agents coordinate through handoffs
- Synthesizer integrates agent outputs

**After**: Reasoning Protocols
- Protocols guide reasoning patterns
- Reasoning streams follow structural templates
- Context flows through intelligent handoffs
- Integration synthesizes reasoning streams

The shift is from **"agents that do"** to **"protocols that guide thinking."**

## Protocol Architecture

### Core Reasoning Protocols

Each protocol defines a structured approach to reasoning about complex problems:

1. **[Hierarchical Reasoning Protocol](./hierarchical-reasoning-protocol.md)**
   - *Purpose*: Tiered reasoning with dependency management
   - *When*: Foundation → Analysis → Strategy problems
   - *Key Concept*: Build understanding in layers, parallelize within tiers
   - *Example*: Market research → Competitive analysis → Market entry strategy

2. **[Sequential Reasoning Protocol](./sequential-reasoning-protocol.md)**
   - *Purpose*: Pipeline thinking with intelligent handoffs
   - *When*: Clear sequential stages with transformations
   - *Key Concept*: Each stage transforms input into refined output
   - *Example*: Requirements → Design → Implementation → Validation

3. **[Strategic Reasoning Protocol](./strategic-reasoning-protocol.md)**
   - *Purpose*: Decision acceleration through constraint exposure
   - *When*: Decision paralysis, authority ambiguity, strategic uncertainty
   - *Key Concept*: Questions over answers, diagnose blockers, map authority
   - *Example*: "Should we enter this market?" → Meta-trait extraction → Stress-test questions

4. **[Competitive Reasoning Protocol](./competitive-reasoning-protocol.md)**
   - *Purpose*: Diverse approaches with merit-based synthesis
   - *When*: Creative problems, uncertain best approach, need validation
   - *Key Concept*: Enforce isolation, evaluate by criteria, synthesize strengths
   - *Example*: Multiple design approaches → Evaluation → Hybrid solution

5. **[Iterative Reasoning Protocol](./iterative-reasoning-protocol.md)**
   - *Purpose*: Progressive refinement toward quality thresholds
   - *When*: Complex reasoning requiring polish, high-stakes thinking
   - *Key Concept*: Measure improvement, recognize diminishing returns
   - *Example*: V1 → V2 → V3 with quality metrics and specific feedback

6. **[Federation Reasoning Protocol](./federation-reasoning-protocol.md)**
   - *Purpose*: Multi-domain integration with domain expertise
   - *When*: Problems spanning distinct domains (technical, financial, legal, market)
   - *Key Concept*: Preserve domain authority, coordinate dependencies, synthesize
   - *Example*: Product launch (technical + market + financial + legal assessment)

7. **[Integration Reasoning Protocol](./integration-reasoning-protocol.md)**
   - *Purpose*: Synthesize multiple reasoning streams coherently
   - *When*: Multiple perspectives need synthesis into unified understanding
   - *Key Concept*: Pattern recognition, conflict resolution, emergent insights
   - *Example*: Multi-stream analysis → Cross-stream patterns → Actionable synthesis

## Protocol Structure

Each protocol follows a consistent structure inspired by the Reasoning Protocol format:

### 1. Role
Clear statement of the protocol's purpose and cognitive responsibility.

### 2. Non-Negotiable Commitments
Core principles that must be upheld when using this protocol.

### 3. Protocol Flow
Step-by-step reasoning process:
- **Step 0**: Assess suitability and context
- **Step 1-N**: Execute protocol-specific reasoning stages
- **Final Step**: Meta-evaluation and learning extraction

### 4. Grammar
Key conceptual distinctions to surface during reasoning.

### 5. Error Forensics
How to diagnose and recover from reasoning failures.

### 6. Cognitive Stance
Mindset and attitude for effective protocol application.

### 7. Internal Checks
Self-validation questions to ensure protocol integrity.

### 8. Ethos
Core philosophy explaining why the protocol works.

### 9. Use Cases
Specific problem types where protocol excels.

### 10. Anti-Patterns
Common mistakes to avoid when applying protocol.

## Protocol Selection Guide

### By Problem Type

**Complex Multi-Stage Problems**
→ Use **Hierarchical** or **Sequential** depending on dependencies
- Hierarchical: When parallel exploration possible within stages
- Sequential: When strict stage-by-stage transformation required

**Strategic Decision Problems**
→ Use **Strategic Reasoning Protocol**
- Decision paralysis signals
- Authority ambiguity
- Risk aversion

**Creative or Uncertain Problems**
→ Use **Competitive Reasoning Protocol**
- Multiple valid approaches
- Benefit from diverse perspectives
- Need validation through comparison

**Quality-Critical Problems**
→ Use **Iterative Reasoning Protocol**
- High-stakes thinking
- Refinement improves quality
- Clear quality criteria exist

**Multi-Domain Problems**
→ Use **Federation Reasoning Protocol**
- Spans technical, financial, legal, market domains
- Requires domain expertise preservation
- Needs cross-domain integration

**Multiple Perspectives to Synthesize**
→ Use **Integration Reasoning Protocol**
- Multiple reasoning streams exist
- Need coherent unified understanding
- Conflicts require resolution

### By Reasoning Structure

**Foundation → Analysis → Strategy**
→ Hierarchical Reasoning Protocol

**Stage 1 → Stage 2 → Stage 3 → ... → Stage N**
→ Sequential Reasoning Protocol

**Parallel Streams → Evaluation → Synthesis**
→ Competitive Reasoning Protocol

**V1 → V2 → V3 (refinement cycles)**
→ Iterative Reasoning Protocol

**Domain 1 + Domain 2 + Domain 3 → Integration**
→ Federation Reasoning Protocol

**Stream A + Stream B + Stream C → Unified Understanding**
→ Integration Reasoning Protocol

### By Primary Goal

**Build understanding in layers**: Hierarchical
**Execute pipeline with handoffs**: Sequential
**Accelerate blocked decisions**: Strategic
**Explore diverse approaches**: Competitive
**Refine toward quality threshold**: Iterative
**Coordinate across domains**: Federation
**Synthesize multiple streams**: Integration

## Protocol Combination

Protocols can be combined for complex reasoning:

### Example: Product Launch Decision

1. **Strategic Protocol** (Is this the right time?)
   - Extract meta-traits
   - Generate stress-test questions
   - Map authority

2. **Federation Protocol** (Multi-domain analysis)
   - Technical feasibility assessment
   - Market opportunity analysis
   - Financial viability assessment
   - Legal compliance review

3. **Integration Protocol** (Synthesize domains)
   - Identify cross-domain patterns
   - Resolve conflicts
   - Create go/no-go framework

4. **Hierarchical Protocol** (If go: plan execution)
   - Tier 1: Research and preparation
   - Tier 2: Development and testing
   - Tier 3: Launch strategy and execution

### Example: High-Stakes Report

1. **Hierarchical Protocol** (Structure research → analysis → recommendations)
   - Tier 1: Parallel research streams
   - Tier 2: Analytical synthesis
   - Tier 3: Strategic recommendations

2. **Iterative Protocol** (Refine quality)
   - V1: Initial draft
   - V2: Clarity and coherence improvements
   - V3: Final polish

3. **Integration Protocol** (Final synthesis)
   - Synthesize all reasoning
   - Create executive summary
   - Ensure actionability

## Key Concepts

### Reasoning vs Execution
**Execution**: Agents perform tasks (write code, gather data, create designs)
**Reasoning**: Protocols guide thinking (analyze, synthesize, decide)

These protocols guide reasoning, not task execution.

### Context Intelligence
All protocols emphasize intelligent context management:
- **Context Building**: What information is needed?
- **Context Filtering**: How should information be curated?
- **Context Framing**: Why is this information relevant?

### Information Flow Design
Each protocol specifies how information flows:
- **Sequential Flow**: Stage-by-stage handoffs
- **Tiered Flow**: Tier-to-tier with filtering
- **Parallel Flow**: Independent streams converging
- **Federated Flow**: Domain-specific to cross-domain

### Quality Over Speed
Protocols emphasize quality of reasoning:
- Clear criteria and thresholds
- Explicit tradeoffs and conflicts
- Source attribution
- Confidence qualifications

### Integration as Value Creation
Integration is not concatenation:
- **Bad**: "Stream A says X, Stream B says Y"
- **Good**: "Across streams, pattern P emerges revealing insight I"

## Implementation Approach

### As Prompts
Use protocols as system prompts for AI models:
```
You are applying the Hierarchical Reasoning Protocol.

[Include full protocol text]

Now reason about: [problem statement]
```

### As Reasoning Templates
Use protocol structure to organize human reasoning:
- Follow step-by-step flow
- Apply grammar distinctions
- Check against anti-patterns

### As Quality Standards
Use protocols to evaluate reasoning quality:
- Does reasoning follow protocol structure?
- Are protocol commitments upheld?
- Are anti-patterns avoided?

### As Cognitive Tools
Use protocols as mental models:
- Recognize when to apply which protocol
- Diagnose reasoning failures using error forensics
- Improve reasoning through protocol mastery

## Migration from Agentic Patterns

### Hierarchical Pattern → Hierarchical Reasoning Protocol
- **Agent Tiers** → **Reasoning Tiers**
- **Agent Handoffs** → **Context Handoffs**
- **Agent Dependencies** → **Reasoning Dependencies**

### Linear Pattern → Sequential Reasoning Protocol
- **Agent Pipeline** → **Reasoning Pipeline**
- **Agent Outputs** → **Stage Outputs**
- **Agent Coordination** → **Context Flow**

### Strategic Delegator → Strategic Reasoning Protocol
- **Agent Team Creation** → **Reasoning Dimension Design**
- **Agent Execution** → **Structured Analysis**
- **Agent Synthesis** → **Framework Creation**

### Competitive Pattern → Competitive Reasoning Protocol
- **Multiple Agents** → **Multiple Approaches**
- **Agent Isolation** → **Stream Isolation**
- **Agent Evaluation** → **Criteria-Based Evaluation**

### Iterative Pattern → Iterative Reasoning Protocol
- **Version Agents** → **Reasoning Versions**
- **Agent Feedback** → **Quality Assessment**
- **Agent Refinement** → **Systematic Improvement**

### Federation Pattern → Federation Reasoning Protocol
- **Domain Agents** → **Domain Reasoning**
- **Agent Coordination** → **Cross-Domain Coordination**
- **Agent Integration** → **Domain Synthesis**

### Synthesizer Agent → Integration Reasoning Protocol
- **Agent Synthesis** → **Stream Integration**
- **Agent Outputs** → **Reasoning Streams**
- **Agent Coordination** → **Pattern Recognition**

## Benefits of Protocol Approach

### Cognitive Clarity
- Explicit reasoning structure
- Clear decision points
- Traceable logic

### Scalability
- Protocols scale to any complexity
- Can be combined and nested
- Work for human and AI reasoning

### Quality Assurance
- Built-in error detection
- Clear quality criteria
- Systematic improvement

### Reusability
- Patterns apply across domains
- Templates accelerate reasoning
- Learn from protocol mastery

### Transparency
- Reasoning process visible
- Assumptions explicit
- Tradeoffs documented

## Next Steps

1. **Study Protocols**: Read and understand each protocol's structure and intent
2. **Practice Application**: Apply protocols to real problems
3. **Recognize Patterns**: Learn when to use which protocol
4. **Combine Protocols**: Master protocol composition for complex reasoning
5. **Refine Protocols**: Adapt and improve based on experience

## Reference

For the original Reasoning Protocol that inspired this structure:
See the query that initiated this transformation.

For the original agentic patterns:
See `/agents/patterns/` directory for agent-based implementations.

---

**Remember**: These protocols guide thinking, not execution. They are frameworks for structured reasoning that create coherent understanding from complex problems.