Safe Agentic Workflows

A hands-on course on structuring AI agent automation — from local harnesses to CI/CD-triggered workflows with guardrails, cost controls, and multi-agent coordination.

Foundations

01 The Three-Layer Architecture Hooks, Commands, and Skills — the mental model behind every agent harness. ~10 min
02 gh-aw Quick Start Install the CLI, add your first agentic workflow, and trigger a run. ~10 min
03 The Security Model Safe outputs, sandboxing, and threat detection — how agents earn trust. ~12 min

Automation Patterns

04 Triggers and Events Making agents react to issues, PRs, comments, and schedules. ~8 min
05 Background Agents Scheduled automation — wake up to ready-to-review improvements. ~10 min
06 Design Patterns IssueOps, ChatOps, MonitorOps, SpecOps — choosing the right pattern. ~10 min

Scaling Up

07 Cross-Provider Portability One harness across Claude Code, Gemini CLI, Codex, and Cursor. ~10 min
08 Cost Controls and Observability AI credit budgets, staged mode, and tracking what agents spend. ~8 min
09 Multi-Agent Coordination Orchestration, subagents, role collapsing — from solo to team. ~12 min
10 Putting It All Together A phased implementation plan — from your existing harness to full automation. ~15 min

Course: Safe Agentic Workflows — 10 lessons on building AI agent automation with guardrails.

Total course time: ~105 minutes · 10 lessons · Interactive quizzes in each

← All Courses