Log an ambiguous or uncertain finding from a code review into a running technical-debt ledger instead of blocking the review on it. Use when a reviewer is unsure whether something is actually wrong ("애매하다", "아리까리하다", "확신이 안 선다", "이건 일단 부채로") and wants to note…
Post zereight-review findings as Bitbucket inline PR comments (simple English body + ASCII diagrams + short Korean summary at end, code fixes). Use after zereight-review when the user asks to post review comments, inline comments, "댓글 달아", or…
PR/diff code review for JS/TS. TypeSafe Jev (typesafe/jev-1.13 via OpenRouter) screens every changed file into a risk matrix and finding candidates, the tool collects each top file's neighbors (importers through barrels, imports, siblings), then the agent…
Comprehensive code review skill for practical PR feedback. Use for feature, bugfix, and refactor reviews. Prioritizes correctness, edge cases, dual-path symptom fixes vs delete-a-path alternatives, logic invariants, async state, flow ownership, ponytail…
Single entry for zereight/gitlab-mcp maintainer work — triage, review, verify, fix, merge; short Korean follow-ups (머지 ㄱ, 수정해줘, ~된거지?); repo-state checks before release/CI claims; explain findings with code, metaphor, or ASCII diagrams. Loads…
Audit a UI or design against WCAG 2.2 AA/AAA and ARIA patterns, returning criterion-referenced findings with severity and specific fixes. Use when the user wants an accessibility check, contrast verification, keyboard/screen-reader review, or wants to confirm…
Apply a visual direction — an archetype (high-end agency, editorial minimal, brutalist, soft-SaaS, dark-tech) or one of 138 named design systems (apple, linear-app, stripe, vercel, notion, material, shadcn, spotify, tesla…) — by resolving it into the token…
Run iterative agent evaluation and improvement loops (judge, improve, scenarios, playbooks) via pi-autocontext or autoctx CLI. Use when improving repeatable agent workflows (review rubrics, skill quality, scenario-based feedback), checking run status, or…