iterative-reviewlisted

Iterative review-fix-verify loop for code, docs, and phasing-run artifacts. Runs review → fix → verify → re-review with severity-gated stopping: Critical always gates the user, Important auto-fixes by default (per-finding flip available), Suggestions surface in the report only. 3-iteration hard cap, skip-if-clean pre-gate (avoids the Snorkel self-critique 41pt accuracy drop), fixed-point early exit, 5× token-budget circuit breaker. Auto-detects mode: code (PR or local diff), doc (.md / .txt / .rst / specs/ / docs/), or phase (phasing-run deliverables vs. spec). In phase mode, plan-affecting findings flow through the existing handoff "Open questions for downstream" channel without modifying the locked master plan. Use when the user says "/iterative-review", "review and fix", "loop the review", "verify phase N output", "review the spec against the PR", or wants more than a one-shot review. Use proactively after `/pr-review-toolkit:review-pr` returns Critical findings or after a phasing phase completes and its d
unifylabs-dev/unify-kit · ★ 0 · Code & Development · score 63

Install: claude install-skill unifylabs-dev/unify-kit

# Iterative Plan-Aware Review Review → fix → verify → re-review loop with bounded iteration. Replaces the brittle "run review again after fixes" pattern with explicit stopping rules. Plan-aware: when running inside a phasing run, plan-affecting findings flow back through the handoff Open-Questions channel without mutating the locked master plan. ## Invocation forms | Form | Meaning | |------|---------| | `/iterative-review` | Auto-detect mode from cwd (see `references/modes.md`) | | `/iterative-review <PR#>` | Code mode against a GitHub PR | | `/iterative-review <path>` | Doc mode if `.md` / `.txt` / `.rst` / `.mdx`; otherwise code mode | | `/iterative-review phase <run-id> <N>` | Phase mode — review phase N's deliverables against its spec | Flags (optional; defaults match recommended behavior): - `--include-suggestions` — surface Suggestion-severity findings in the loop (default: report-only) - `--gate-important` — gate every Important finding instead of auto-fixing - `--cap N` — override the 3-iteration cap (max 5) - `--no-skip-clean` — disable the skip-if-clean pre-gate (NOT recommended — see Step 4) ## Architecture ``` detect mode → detect verifier → initial review pass → skip-if-clean? → EXIT clean → loop (max 3): categorize findings → GATE Critical → AUTO Important → dispatch fix subagents → run verifier → re-review (delta scope only) → fixed-point check → EXIT if stalled → no Critical left → EXIT clean → cap hit → EXIT wi