Skip to content
Agentic Engineering Ebook
Search
Ctrl
K
Cancel
Kiberon Labs
Select theme
Dark
Light
Auto
Cover
Preface
Introduction
Part I: One Agent
Introduction
What an Agent Is
Models
Prompting
Tools
Agent Loop
Part II: A Reliable Agent
The Harness
Agent-Computer Interface
Verification & the Outer Loop
Controls & Autonomy
Security & Prompt Injection
Deep dives
Guardrails
ACRI
Foundational AI Model
MCP
Part III: What the Agent Knows
Context
The Knowledge Base
Skills
Agent Readiness
Deep dives
Context Topics
Context Layers
Runtime Context Management
Context Development Lifecycle
Context Evaluation
Knowledge Base Topics
Retrieval-Reasoning Spectrum
RAG-RLM Hybrid Architecture
Routing Heuristics
Adoption Sequence
Skill Format
Skill Gateway
Skill Gateway Prior Art
Part IV: Delivering Real Work
Agentic Software Development
Ticket Enrichment
Planning & Execution
End-to-End Orchestration
Deep dives
Execution
Orchestration Triggers
Model Usage
Smart Routing
Collaborative Development Structure
Economics & Routing
Execution & Feedback
Operations & Governance
Part V: Many Agents
The Agentic Swarm
The Coordination Model
Agent-to-Agent Protocols
Agent
Agent Pool
Cohort
Orchestrator
Agent Architecture
Swarm Operations
Deep dives
The Execution Ledger
Part VI: The Self-Improving System
The Evaluation Harness
Skill Evaluation
Ikidna in Practice
Agent Container Runtime Interface (ACRI)
Architecture
Axioms
Goals
Harnesses
Minimal Bootstrap
Bootstrap
Minimal Goals
Prior Art & Lessons
Kiberon Labs
Select theme
Dark
Light
Auto
Current
Agent Harness
Section titled “Agent Harness”
Current
Section titled “Current”
Opencode
Potential
Section titled “Potential”
Forgecode
Pi
Context / Knowledge Engine
Section titled “Context / Knowledge Engine”
Potential
Section titled “Potential”
GitNexus
Mem0
Graphiti / Zep
Evaluation / Skills Context
Section titled “Evaluation / Skills Context”
Tessl
Promptfoo
