메뉴
Investigate whether an AI coding change matches the user's task and whether cross-module changes are necessary. Use only for an explicit /aet-scope request.
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
Create a bounded, read-only investigation from normalized Codex or Claude Code Run Records, optionally inspect explicit AET Proof and Freshness records, and export a Portable Evidence Bundle for an independent reviewer.
Run a bounded AET check of Agent instructions, Skills, verification contracts, and dangerous authority. Use only when the user explicitly invokes /aet-check or asks for this AET Quick check.
Deterministically check whether an AET proof still applies to the current workspace. Use only for an explicit /aet-fresh request.
Execute one explicit verification command and record a minimal hash-bound AET proof receipt. Use only when the user explicitly invokes /aet-proof or asks AET to prove a command ran.
Compatibility entrypoint for AET Quick and advanced AET Lab workflows. Prefer the four dedicated Quick Skills for daily work; use this Skill only when the user explicitly asks for legacy CLI or Lab capabilities.
Held-out fixture with a mismatched directory name.
| name | aet-scope |
| description | Investigate whether an AI coding change matches the user's task and whether cross-module changes are necessary. Use only for an explicit /aet-scope request. |
Answer one question: does each change group fit the user's authorized task?
Run aet quick scope <root> --base <revision> --intent <intent.json> --format json for the deterministic preflight. Recover intent in this order:
current request, later authorization, explicit Intent, Issue/PR, Commit
message, then clearly marked inference.
A path outside a declared surface is only a fact that triggers investigation.
It is never sufficient by itself for OUT_OF_SCOPE.
Group changes by purpose. For each material group, examine direct dependencies, shared infrastructure, interface migration, tests, generated files, later authorization, and a reasonable counter-hypothesis. Store each useful tool step in an Investigation Ledger with its question, immutable result reference, observation, hypothesis effect, decision value, and cost.
Allowed dispositions are IN_SCOPE, JUSTIFIED_EXPANSION,
POSSIBLE_SCOPE_EXPANSION, OUT_OF_SCOPE, and INSUFFICIENT_INTENT.
Before rendering an Investigated Finding, pass the canonical Finding, Ledger,
authorized-tool set, and Investigation Contract through
aet.investigation.validate_investigated_finding. If that API is unavailable
or rejects the record, preserve HYPOTHESIS/UNKNOWN and disclose why.
Budget: 45 seconds, four rounds, two LLM calls, eight tool calls, two authorized remote reads, one expensive call, and at most one low-cost discriminating test. No code writes.
For a Chinese /aet-scope request, use natural Simplified Chinese while
keeping code and required technical terms in English. Otherwise use English.
Report facts, judgment, counter-explanation, uncertainty, location, and the smallest next action, then stop. Do not repair, prove, publish, or enter Lab.
