5 tagged git hooks, measured the same way as everything else here.
Agent Claude Code
Bounded Opus build leg for mechanical labor against a ruled spec — fixtures, mechanical variations on an exemplar, sweeps, file authoring to a verbatim contract clause. Never handed a ruling or contract judgment.
Agent Claude Code
Read-only-charge Opus exploration/derivation leg — censuses, sweeps, searches, state derivation across many files or repos. Returns grounded conclusions with named sources; no-edit tool grant.
Agent Claude Code
Standing pre-merge falsification leg (mmnto-ai/totem-strategy:doctrine/model-tiering.md § Review legs). Dispatch on any self-authored judgment-dense diff BEFORE presenting it for merge, and again on a fold that rewires semantics (re-armed, scoped to the fold diff). Reviews falsification-first against named primary…
Agent
Clean-context design reviewer. Reviews just-written code changes with zero memory of writing them — catches half-finished changes and cross-file drift (a file updated but the file pointing at it left stale), then over-engineering, scope creep, and simpler alternatives: the judgment calls hooks and checklists cannot…
Agent
Evidence verifier for repo-xray scan results. Opens the actual code at each scanner candidate and decides which ones are real — filtering the noise a regex/AST scan inevitably produces. Use for /hi-vibe:check (spawned by default), or when the user asks 이거 진짜 중복이야?, 스캔 결과 확인해줘, 후보 걸러줘. Not for design opinions or…