Learn agentic AI at your own pace, one short lesson at a time.

Design a Software Factory for Your Coding Agents

1. What a Software Factory AutomatesFree
2. Stop Babysitting Your Coding AgentPro
3. Verification Is the BottleneckPro
4. Turn Agent Rules Into GatesPro
5. Boundary, Context and SkillsPro
6. Isolate Every Agent RunPro
7. Why Verification Is the Hard LayerPro
8. Tests First, Then Verify Outside the AgentPro
9. From Pull Request to MergePro
10. Human Review Stays in the LoopPro

Build Your First Software Factory Execution Harness

1. Six Layers Around the Execution LayerFree
2. Isolate Ports, Databases and EnvironmentsPro
3. Run Coding Agents in the BackgroundPro
4. Stop Parallel Agents Sharing One BranchPro

Build a Claude Code Verification Harness

1. Six Layers of a Software FactoryFree
2. CLAUDE.md Is Guidance, Not VerificationPro
3. Why Claude Code Cannot Grade ItselfPro
4. Rules Ask, Gates CheckPro
5. Deny Rules and Hooks That RefusePro
6. The Stop Hook Closes the LoopPro
7. Will This Process Outlast the Next Model?Pro

Build an Autonomous Claude Code Agent

1. A Software Factory Is Not a Dark FactoryFree
2. Boundary and Context Are YoursPro
3. Skills Stop the Agent ImprovisingPro
4. Verify Outside the AgentPro
5. The Repo Carries the HarnessPro
6. Run Claude Headless on an IssuePro
7. Hooks Refuse Before the RunPro
8. The Stop Hook Sends Work BackPro
9. What It Takes to Automate a PRPro

Build a Mission Control for AI Agents

1. Verify What Your AI Agent Calls DoneFree
2. An AI Agent Team With a Mission Control BoardPro
3. Human Approval Belongs in Code, Not the PromptPro
4. Done Is Decided by a Check, Not the AgentPro
5. Cap AI Agent Loops With Cost and Max TurnsPro

Build a Multi-Agent System with MCP

1. When to use one AI agent or a team of agentsFree
2. What happens inside an AI agent loop, and where MCP fitsPro
3. Give each AI agent one job and only its toolsPro
4. Your AI agent needs the same auth as a personPro
5. Ship an AI agent like an API with timeouts and tracesPro
6. Long-running AI agents should wait, not runPro

Secure AI Agents with Guardrails

1. Why AI agents need different guardrails than APIsFree
2. Stop prompt injection by taking away capabilityPro
3. Keep a human in the loop when agents change statePro
4. Make AI agents answer from your data, not their ownPro
5. Limit what an AI agent can reach: tenants, schemas, PIIPro
6. Enforce AI agent permissions in code, not the promptPro
7. Cap AI agent loops and move defenses out of the modelPro

Lessons out soon