The central navigation and orientation layer for teams adopting agentic AI practices — connected to three companion repos covering skills & knowledge, twin runtime deployment, and business ROI methodology.
Not sure where to start? → audiences/README.md
| Audience | Start Here |
|---|---|
| 👷 Engineers — building agentic features into codebases | audiences/engineers.md |
| 🏗️ Architects — designing multi-agent systems and governance | audiences/architects.md |
| 📈 Data & Analysts — measuring AI impact with ROI frameworks | audiences/data-analysts.md |
| 🎯 Executives — evaluating AI adoption strategy and business case | audiences/executives.md |
| ⚡ Power Users — building and deploying digital twins end-to-end | audiences/power-users.md |
| 🎓 AI Implementation Fellows — OI Lab 12-week program participants | audiences/fellows.md |
| 🔒 Compliance & Risk Officers — governance, HITL, and audit trails | audiences/compliance-risk.md |
| 📋 Product Managers — scoring AI use cases and tracking ROI | audiences/product-managers.md |
This playbook connects three companion repositories. Each is a focused tool; the playbook tells you when and how to use them.
graph TD
PB["📚 agentic-playbook\nNavigation & Orientation"]
AK["🧠 agent-kernel\ngithub.com/fszale/agent-kernel"]
AF["🏭 agent-factory\ngithub.com/fszale/agent-factory"]
OI["📊 operational-intelligence-lab\ngithub.com/fszale/operational-intelligence-lab"]
PB -->|"22 skills · 16 prompts · 12 templates"| AK
PB -->|"twin runtime · API · Cloud Run"| AF
PB -->|"ROI framework · curriculum · cohorts"| OI
style PB fill:#1a1a2e,stroke:#e94560,color:#fff
style AK fill:#0f3460,stroke:#e94560,color:#fff
style AF fill:#0f3460,stroke:#e94560,color:#fff
style OI fill:#0f3460,stroke:#e94560,color:#fff
| Repo | What It Is | 1-Pager |
|---|---|---|
| agent-kernel | Portable agent knowledge layer: 22 skills, 16 prompts, 12 templates, 9 diagrams | ecosystem/agent-kernel.md |
| agent-factory | FastAPI runtime for hosting digital twins with Supabase, admin UI, Cloud Run | ecosystem/agent-factory.md |
| operational-intelligence-lab | Business AI deployment playbook: ROI methodology, dual-track curriculum, cohorts | ecosystem/operational-intelligence-lab.md |
Conceptual introductions to agentic practices. Start here if you are new.
| Guide | Description |
|---|---|
| Antigravity Primer | The injectable knowledge model and why it matters |
| Environment Selection Guide | Antigravity vs. Copilot vs. Claude vs. Codex |
| Implementation, Execution & Gap Analysis | The three mandatory phases of agentic work |
| Building Agentic Skills | From disposable prompts to persistent team assets |
| Analysis Gap & Traceability | Ensuring agents build what you actually need |
| LLM Models Comparison | Claude · Gemini · OpenAI · Grok — when to use each |
Architectural and implementation patterns for production-grade agentic systems.
| Guide | Description |
|---|---|
| Agentic Factories | Multi-agent pipelines with strict data contracts |
| Memory and Context | Short-term, long-term, and working memory |
| Self-Improvement Loop | Systems that get smarter from their own failures |
| HITL & Guardrails | Human-In-The-Loop approvals and the autonomy ladder |
| Digital Twin Agent | Building a deployable expert-encoded AI proxy |
| File | What It Contains |
|---|---|
| PHILOSOPHY.md | Four-lens decision framework |
| GLOSSARY.md | Plain-language definitions for key terms |
| CONTRIBUTING.md | How to add guides, audiences, and ecosystem summaries |
| CHANGELOG.md | Content history |
| CONTEXT.md | AI-first project map (for agents) |
| AGENTS.md | Agent navigation rules (for agents) |
Read CONTEXT.md first, then AGENTS.md.