메뉴
Measure cost and context before justifying any harness change (P7). Read /context, /cost, /status and the tool-failure log to get before/after numbers; roll back anything that worsens a metric. Use when adding tools, evaluating a phase, or diagnosing quota burn.
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
GitNexus is a conditional accelerator, not a default navigation tool. Use it early only when the task already signals large-repo, cross-module, blast-radius, or unclear-flow uncertainty. Every candidate must be verified with rg/Probe before reading deeply or editing.
Offload code search to deterministic tools instead of hand-grepping or reading whole files (P1). Climb the ladder lexical → structural → semantic, stopping at the first rung that answers. Use whenever locating symbols, call sites, definitions, or concepts in a repo.
Compress output to the minimum that conveys the result (P8). Drop preamble, prose, and whole-file re-prints; deliver code, diffs, and file:line refs. Use on every response, scaling density to task size (lite / full / ultra).
Offload mechanical and type-aware edits to deterministic tools instead of LLM file-rewrites (P2). Use ast-grep --rewrite for repeated/mechanical transforms and Serena replace_symbol for single type-aware symbol edits. Apply whenever a change is rule-shaped rather than judgment-shaped.
Engineering guardrails for every change — the behavioral base of the policy spine. Make no unrequested assumptions, no over-engineering, no orthogonal edits, and match the surrounding code. Apply on every edit, plan, and review.
Portable code-review checklist. Use before a commit/PR or after a non-trivial change. Run it in a separate read-only helper when your runtime supports one; Sober ships the checklist, not a fixed reviewer.
| name | observe |
| description | Measure cost and context before justifying any harness change (P7). Read /context, /cost, /status and the tool-failure log to get before/after numbers; roll back anything that worsens a metric. Use when adding tools, evaluating a phase, or diagnosing quota burn. |
No addition is justified by intuition. Every tool, skill, or hook must pay for itself in a measured metric; if a change worsens a number, revert it (P7).
| Metric | Source |
|---|---|
| Files read / task | session transcript count |
| Output tokens | /cost |
| Peak context fill % | /context |
| Message count | /cost, /status |
| Turn count | session |
| Retry rate | tool-failure log |
| Failure blast radius | files touched per failed change |
/context — token budget + peak fill./cost, /status — messages, output tokens.scripts/hooks/tool-failure-log.sh output — retries, failure patterns.One change at a time, measured. Bundled changes hide which one regressed. Keep a short before/after note in HANDOFF.md (P5), not auto-injected.
Default memory is file-based (HANDOFF.md + .serena/memories). Only if a metric shows it's insufficient: claude-mem (auto session memory) — watch for auto-injection noise. Adopt via the before/after gate above, never preemptively (ADR-010/P7).