Skip to Content
Nextra 2 Alpha
ExamplesFrom Implicit to Explicit Knowledge

From Implicit to Explicit Knowledge

This example demonstrates how Rhema transforms scattered, ephemeral knowledge into structured, persistent context that survives across AI conversations and development sessions.

🎯 The Problem: Lost Context

In traditional development workflows, critical knowledge exists in:

  • Individual minds - Developer memories and experiences

  • Temporary chats - AI conversations that disappear

  • Scattered docs - Unstructured, stale documentation

  • Forgotten decisions - Architectural choices with lost rationale

This creates knowledge silos, session amnesia, and inconsistent AI behavior across your team.

💡 The Rhema Solution

Rhema transforms ephemeral knowledge into persistent, structured context that:

  • 🔄 Survives sessions - Context persists across AI conversations and development sessions

  • 👥 Scales with teams - Knowledge is shared and discoverable across your organization

  • 📈 Evolves with code - Context changes are tracked alongside code in Git

  • 🎯 Enables consistency - AI agents access the same structured context

  • ⚡ Accelerates onboarding - New team members quickly understand project context

📋 Before Rhema (Implicit Knowledge)

Developer A: "I remember we decided to use PostgreSQL for the user service..."
Developer B: "Wait, when was that decided? I thought we were using MongoDB."
AI Agent: "Based on the code I can see, I recommend using Redis for caching..."
Developer C: "Actually, we already tried Redis and it caused issues with our deployment."

Problems:

  • ❌ Session amnesia - AI agents don’t remember past decisions

  • ❌ Team misalignment - Different developers have different understandings

  • ❌ Lost knowledge - Critical insights disappear when team members change

  • ❌ Inconsistent recommendations - AI agents make conflicting suggestions

✨ After Rhema (Explicit Knowledge)

Decisions File

# .rhema/decisions.yaml
 
 
decisions:
 
  - id: "decision-001"
    title: "Use PostgreSQL for user service"
    description: "Chosen for ACID compliance and existing team expertise"
    status: "approved"
    date: "2024-01-15"
    rationale: "MongoDB lacks ACID transactions needed for user data integrity"
    alternatives_considered: ["MongoDB", "MySQL"]
    impact: "Affects user-service, auth-service, and payment-service"

Knowledge File

# .rhema/knowledge.yaml
 
 
insights:
  performance:
 
    - finding: "Redis caching caused deployment issues"
      impact: "Service startup failures in containerized environment"
      solution: "Use in-memory caching with periodic persistence"
      confidence: "high"
      evidence: ["Deployment logs", "Performance metrics"]
      related_files: ["src/cache.rs", "docker-compose.yml"]

🎉 Results

  • ✅ Session continuity - AI agents know about the PostgreSQL decision and Redis issues

  • ✅ Team alignment - Everyone has access to the same explicit context

  • ✅ Faster onboarding - New developers can quickly understand past decisions

  • ✅ Consistent recommendations - AI agents make recommendations based on explicit knowledge

  • ✅ Knowledge preservation - Critical insights aren’t lost when team members change

🔍 Querying the Context

Once knowledge is explicit, you can query it:

# Find all approved decisions
 
 
rhema query "decisions WHERE status='approved'"
 
# Find performance insights
 
 
rhema query "knowledge.insights.performance"
 
# Find decisions affecting multiple services
 
 
rhema query "decisions WHERE impact CONTAINS 'multiple'"

🚀 Next Steps

  • Start small - Begin by recording your next architectural decision

  • Build incrementally - Add knowledge as you discover insights

  • Query regularly - Use CQL to find and leverage existing context

  • Share with team - Commit context files to version control

🔗 Related Examples

Last updated on
Enhanced Context Injection SystemLock File Operations