메뉴
Repo-wide or subsystem-wide architecture hygiene pass that surfaces deepening opportunities, shallow modules, leaky seams, and ownership drift. Default mode is report-only: no bead creation or execution unless explicitly requested.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | architecture-rescue |
| description | Repo-wide or subsystem-wide architecture hygiene pass that surfaces deepening opportunities, shallow modules, leaky seams, and ownership drift. Default mode is report-only: no bead creation or execution unless explicitly requested. |
| metadata | {"version":"1.0","ecosystem":"pulse","dependencies":[]} |
Use this skill when the user wants architecture cleanup that is broader than a single diff: repo-wide, subsystem-wide, or cross-module shape correction.
This skill is for analysis and proposal first. By default it is report-only.
Default to report-only unless the user explicitly asks for planning/execution follow-through.
In report-only mode:
pulse:planningpulse:executingFor the target scope (whole repo or named subsystem), read:
CLAUDE.md, AGENTS.md, and nearby module docs)history/<feature>/CONTEXT.md decisions when present.pulse/memory/* entries that apply to architecture constraints, incidents, or non-regression rulespulse:gitnexus and GitNexus context if availableIf GitNexus is configured, prefer it for topology, process flow, and impact/breadth evidence before manual grep.
Use consistent terms from references/LANGUAGE.md.
Look specifically for:
Use the deletion test: if deleting a module merely moves complexity to callers, it was likely earning leverage; if it removes no real complexity, it was likely shallow or pass-through.
Generate a short list (3-7) of high-signal candidates with concrete evidence.
For each candidate capture:
Return a ranked architecture rescue report:
Do not generate implementation tasks unless asked.
Use only when the user explicitly requests them.
phase-shaping: hand off top approved candidate(s) to pulse:planningexecution-ready: after validating gate approval, proceed via pulse:swarming/pulse:executingBefore leaving report-only mode, restate the mode change and requested scope.
Provide concise output with:
Keep recommendations specific, evidence-backed, and bounded to the requested scope.
Use after completed Pulse work when the user asks to capture post-cycle machine-readable learnings for future planning and execution.
Use when the user asks to distill accumulated dev-notes into reader-facing topic knowledge, rather than to capture a new note or update machine memory.
Use when implementing approved Pulse beads, either as a swarm worker or as a single worker when swarming is unavailable.
Use before implementation planning when feature intent is clear but implementation decisions are still unresolved, and the user needs those decisions locked first.
Use after pulse:exploring when locked decisions are ready and the user needs a mode-gated plan with an approved work shape and current-work prep for validation.
Use when starting any Pulse workflow, resuming a Pulse session, or before planning or execution when tool readiness or onboarding state may block the flow.