mit einem Klick
crumb-deep-dive
>-
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Menü
>-
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Recommend a Crumb preset (ambient / video-research / mock) based on the user's setup intent. Trigger on "preset 추천", "추천해줘", "어떤 셋업", "which preset", "suggest a config", "혼자 돌릴까", or hints about installed tools / API keys. Output: recommended preset + reason + alternatives. Default recommendation = ambient (no --preset, entry host follows every actor). Crumb suggests only — user picks. Read-only. Do NOT trigger for game pitches (use `crumb` skill) or status checks (use `crumb-status` skill).
>-
Audit a feature/component for defects by cross-referencing with mature OSS patterns. Produce a prioritized matrix of defects + improvement proposals BEFORE implementing. Use when user says "결함 더 파고들어", "audit this", "find more defects", "improvement proposals first", "이 기능에서 뭐가 잘못됐는지 찾아".
Auto-search and summarize latest AI/agent papers. Triggers: 'paper', '논문', 'arxiv', 'research', '연구', '최신 연구', '학회'.
CHANGELOG.md management rules — Keep a Changelog 1.1.0 + SemVer 2.0.0. Version releases, change logging, scope determination, format standards. Triggers on "changelog", "release", "version", "릴리스", "변경사항", "버전".
Scan the Obsidian wiki and automatically discover missing cross-references between pages. Use this skill when the user says "link my pages", "find missing links", "cross-reference", "connect my wiki", "add wikilinks", "what pages should be linked", or after any large ingestion to ensure new pages are woven into the existing knowledge graph. Also trigger when the user mentions "orphan pages" in the context of wanting to connect them, or says things like "my wiki feels disconnected" or "pages aren't linked well". This is a write-heavy skill — it actually modifies pages to add links, unlike wiki-lint which just reports issues.
Basierend auf der SOC-Berufsklassifikation
| name | crumb-deep-dive |
| description | >- |
Origin: 2026-05-05 사용자 directive — "추정하지 말고 로그 혹은 이벤트와 같은 관측성 데이터로 원인 확정해" + "이 절차를 스캐폴드로 굳히고". Live smoke 01KQVSE9PC에서 Studio Output blackbox 진단을 Playwright headless probe로 정확히 잡은 것이 first canonical example.
npm run verify (lint:all + typecheck + format:check +
zod:check-envelope + check-lockfile + test + build).npm run sandwich:update-hashes.node scripts/check-lockfile-platform-deps.mjs
(macOS @emnapi pruning 자주 재발).선택 가능한 verification surface 5가지 — 항상 ≥1개 사용:
~/.crumb/projects/<id>/sessions/<ulid>/transcript.jsonl
직접 jq로 query. event kinds / metadata / scores 등 1차 source.scripts/diag-output.mjs 패턴 (있으면 재사용,
없으면 ad-hoc 작성). page.goto + console + pageerror + network 캡처 + screenshot.curl -s http://127.0.0.1:7321/api/{health,doctor, sessions} 또는 /api/sessions/:id/.... server-side state truth.<sessionDir>/agent-workspace/<actor>/spawn-*.log.
subprocess의 raw stdout/stderr.하지 말 것:
해야 할 것:
Format (≤500 words, markdown):
## 관측성 데이터로 확정 — <증상>
### Evidence
- <source 1>: <data citation> → <fact 1>
- <source 2>: <data citation> → <fact 2>
- ...
### Root cause
<file:line + 1-line 인과 chain>
### Implication
<system invariants 위배 여부 / Karpathy P# 위배 여부 / 기존 PR과의
관계 — 가설 부정된 PR이 있으면 명시>
### Improvement options
| 옵션 | LOC | trade-off |
| A | ~N | ... |
| B | ~M | ... |
| 추천 | ★ <옵션> + 이유 |
필수 요소:
그 외 모든 feature / bug fix: 4-stage 강제.
scripts/diag-output.mjs 작성. Playwright headless 9 modules
로드 + console + pageerror + canvas 캡처. 결과: 모든 200 OK / pageerror = "Identifier 'remaining' has already been declared" 1건 / canvas 미생성.artifacts/game/src/scenes/GameScene.js: 231 + :239 const remaining 중복. 5개 추정 옵션 중 4개 데이터로 폐기 (CDN
200 OK / WebGL warning만 / sandbox attr 무관 / canvas 0이 아니라 미생성).
Root cause = builder LLM 작성 실수. Improvement 옵션 A (손수정), B (재빌드),
C (qa-runner spec-driven boot predicate으로 미리 잡기 — PR #231 적용됨,
follow-up).metadata.provider 별 event count + tokens.
결과: google 5 events / step.research.video에 27,902 tokens_in stamped /
agent.stop에는 null.ServiceMap.tsx:154 computeProviderMix가 agent.stop 만 sum
하는 게 root cause. gemini-sdk adapter도 SpawnResult.usage 미반환. 두 곳 fix.자주 쓰는 jq pattern (transcript inspection):
TX=~/.crumb/projects/<id>/sessions/<ulid>/transcript.jsonl
# Per-kind count
jq -r '.kind' "$TX" | sort | uniq -c | sort -rn
# Per-provider events
jq -r '.metadata.provider // "(none)"' "$TX" | sort | uniq -c
# Last N events with from/kind/body
tail -10 "$TX" | jq -r '"\(.ts | split("T")[1] | split(".")[0]) [\(.from)] \(.kind) | \(.body // "" | .[0:120])"'
# Find events with token usage
jq -r 'select(.metadata.tokens_in or .metadata.tokens_out) | {kind, from, in: .metadata.tokens_in, out: .metadata.tokens_out}' "$TX"
# Verifier verdict trail
jq -r 'select(.kind=="judge.score" or .kind=="verify.result" or .kind=="handoff.rollback") | {ts, body: (.body | .[0:150])}' "$TX"
자주 쓰는 Playwright probe (scripts/diag-output.mjs): page load + console +
pageerror + network + canvas state + screenshot.
.skills/anti-deception-checklist/SKILL.md (gitignored) — Stage 3에서
"fake success" / "test deletion" 등 anti-pattern 회피 체크리스트.skills/agent-ops-debugging/SKILL.md — multi-agent system root cause
분석 패턴 (Safe Default anti-pattern, ContextVar DI 등)wiki/synthesis/ — Stage 3 보고가 wiki page로 promote될 가치가 있을 때
작성 위치