Practical patterns, starters & CLI tools for loop engineering with AI coding agents. Design systems that prompt and orchestrate agents (inspired by Addy Osmani and Boris Cherny). Includes loop-audit, loop-init, loop-cost.
TypeScript agentic-aiai-agentsai-codinganthropicautomationclaudeclaude-codecodexcoding-agentsdevops-automationdevtoolsgithub-actionsgrokllmloop-engineeringmcpprompt-engineering
chore(daily-triage): add post-run critique section (#78)
Cobus Greyling2e0d0c8ec3d2
diff --git a/patterns/daily-triage.md b/patterns/daily-triage.mdindex a3cf364..d20e86a 100644--- a/patterns/daily-triage.md+++ b/patterns/daily-triage.md@@ -50,6 +50,17 @@ Fields the loop must update every run:4. (Phase 2) For small, self-contained failures: open worktree → implementer → verifier.5. (Phase 3) Connectors update PRs/tickets; ambiguous items flagged for human.6. Prune resolved/merged items from state.+7. Record post-run critique in state: false positives, repeated items, re-prioritized or dropped items, and one adjustment for next run.++## Post-Run Critique++After each Daily Triage run, record:++- High-noise items.+- False positives (items incorrectly flagged).+- Items that should be deprioritized.+- Any human-review friction.+- One change to improve the next cycle.## Verification Strategy@@ -90,6 +101,7 @@ Fields the loop must update every run:| State file grows unbounded | Prune merged/closed items every run || Auto-fix on wrong priority | Start report-only; add explicit effort/risk gates || Missed overnight failures | Add `fireImmediately: true` or run at start of day + mid-day |+| Stale critique / never reviewed | Add human handoff when critique entries accumulate without resolution across N runs. |## Cost Profilediff --git a/starters/minimal-loop/STATE.md.example b/starters/minimal-loop/STATE.md.exampleindex 1168b1e..d61417d 100644--- a/starters/minimal-loop/STATE.md.example+++ b/starters/minimal-loop/STATE.md.example@@ -8,5 +8,12 @@ Last run: never## Recent Noise (ignored this run)+## Post-Run Critique (from last run)+- High-noise: dependabot PRs surfaced again — add to ignore list+- False positives: 1 CI flake (known flaky test)+- Deprioritize: lint warnings moved to Watch List+- Friction: triage missed nightly deploy failure (was infra, not code)+- Adjustment: include infra check status in scan+---Run log: —\ No newline at end of file