Agentic Engineering
Engineering workflows where AI agents perform most implementation and humans enforce quality and risk controls.
When to Activate
- Setting up AI-assisted development workflows
- Planning agent task decomposition
- Optimizing model selection for cost/quality
- Reviewing AI-generated code
Operating Principles
- Define completion criteria before execution
- Decompose work into agent-sized units
- Route model tiers by task complexity
- Measure with evals and regression checks
Eval-First Loop
- Define capability eval and regression eval
- Run baseline and capture failure signatures
- Execute implementation
- Re-run evals and compare deltas
Task Decomposition
Apply the 15-minute unit rule:
- Each unit independently verifiable
- Each unit has a single dominant risk
- Each unit exposes a clear done condition
Model Routing
| Tier | Use For | Cost | |------|---------|------| | Haiku | Classification, boilerplate, narrow edits | Lowest | | Sonnet | Implementation, refactors | Medium | | Opus | Architecture, root-cause analysis, multi-file invariants | Highest |
Escalate model tier only when lower tier fails with a clear reasoning gap.
Session Strategy
- Continue session for closely-coupled units
- Start fresh session after major phase transitions
- Compact after milestone completion, not during active debugging
Review Focus for AI-Generated Code
Prioritize:
- Invariants and edge cases
- Error boundaries
- Security and auth assumptions
- Hidden coupling and rollout risk
Skip: Style-only disagreements when automated format/lint already enforces style.
Cost Discipline
Track per task:
- Model used
- Token estimate
- Retries
- Wall-clock time
- Success/failure