Borrowing it
Nothing to install: this file belongs to claude-code-expert/carve-harness. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/claude-code-expert/carve-harness/main/.claude/skills/eval-init/SKILL.mdgit clone --depth 1 https://github.com/claude-code-expert/carve-harnessWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/claude-code-expert/carve-harness/eval-init)<a href="https://agentmods.dev/skills/claude-code-expert/carve-harness/eval-init"><img src="https://agentmods.dev/badge/skills/claude-code-expert/carve-harness/eval-init/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/claude-code-expert/carve-harness/eval-init"><img src="https://agentmods.dev/badge/skills/claude-code-expert/carve-harness/eval-init.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Agent Snooping · line 161 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00108 | $0.03524 |
| Opus 5.5 | $0.00043 | $0.01410 |
| Sonnet 5.5 | $0.00022 | $0.00705 |
| Haiku 4.5 | $0.00011 | $0.00352 |
Grade A, and why
eval-init scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 165 lines — stays where its author put it; the contents beside it link to each section on GitHub.
eval-init — 평가·품질 게이트 확정 인터뷰
/eval은 골든셋(specs/goldenset/*.json)이 있어야 의미가 있는데, 설치 직후엔 비어 있다.
이 스킬이 그 빈칸을 분석 + 대화형 확정 + 궤적 검사로 채운다. 채점 절차의 정본은
eval-goldenset 스킬이고, 이 스킬은 그 절차를 실행하는 쪽이다
(checklist-loop(SOP) ⟷ carve-verify-loop(실행)와 같은 구조).
--dry-run이면 S4 초안 제시까지만 하고 파일을 쓰지 않는다.
절대 규칙 (어기면 골든셋이 죽는다)
- 자동 확정 금지. 에이전트가 혼자 만든 케이스는 자기가 이미 통과하는 것만 담는다(자기강화). 크리티컬 경로·실패 소재·엄격도는 반드시 사람이 정한다.
- 궤적 검사를 통과한 케이스만 편입한다. 미승인 케이스는 파일에 쓰지 않는다.
- 상태 assert 우선. "했다"는 텍스트가 아니라 파일·명령 exit로 채점한다.
- 기존 자산 불가침. 골든셋이 이미 있으면 덮어쓰지 않고 증설만 한다.
- 질문은 한 번에 하나씩, 근거(분석 결과)를 함께 제시한다. 추측으로 답을 대신 채우지 마라.
S0. 사전 점검 (질문 없음)
- 하네스가 갖춰졌는지 확인. 둘 중 하나라도 만족하면 진행한다:
- 설치본:
.claude/harness-manifest.txt존재 - 하네스 소스 레포:
install.sh+.claude/hooks/+VERSION존재(manifest는 설치본에만 생긴다) 둘 다 아니면 중단 — "설치된 하네스가 아닙니다. 먼저bash install.sh".
- 설치본:
git status --porcelain— dirty면 경고만("결과를 git diff로 검토·롤백하려면 clean 권장").specs/goldenset/*.json존재 → 증설 모드로 전환하고 그렇게 고지한다(기존 케이스·suite 이름 보존).specs/eval-score.json존재 → 기존 baseline 점수를 보여준다(S3 Q6의 근거).
S1. 프로젝트 분석 (질문 없음 — 근거와 함께 1문단 요약)
읽기 전용으로만 수집한다. 쓰지 마라.
| 축 | 방법 | 쓰임 |
|---|---|---|
| 스택·테스트 러너·커버리지 도구 | 설치 팩(.claude/harness-packs)의 .claude/stacks/<pack>.sh에서 STACK_TEST_CMD_HINT·STACK_COVERAGE_MIN을 읽는다. 팩 없으면 package.json·pyproject.toml·go.mod·build.gradle 직접 감지 |
케이스의 cmd_exit0 형태 · Q4 기본값 |
| 진입점 목록 | 라우트·핸들러·CLI 서브커맨드·public API 스캔 | 크리티컬 경로 후보 |
| 수정 빈도 상위 | git log --since=3.months --name-only --pretty=format: | sort | uniq -c | sort -rn | head -15 |
자주 깨지는 곳 = 1순위 후보 |
| 하네스 차단 이력 | cat logs/*.jsonl | jq -r 'select(.decision=="block") | [.tool,.reason//"-"] | @tsv' | sort | uniq -c | sort -rn |
실제 막아낸 실수 패턴 |
| 기존 테스트·커버리지 | 테스트 파일 수, 가능하면 커버리지 실측 1회 | Q4 제안의 근거 |
| CI | .github/workflows/ |
Q7 선택지 |
요약은 숫자와 파일 경로를 포함해 한 문단으로 보여준다. 근거 없는 단정 금지. 분석 결과가 비면(진입점 0·git 이력 없음) 그 사실을 말하고, 합성 케이스로 갈지 물어라.
S2. 인터뷰 A — 무엇이 중요한가 (사람만 아는 것)
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago Changed · +9 lines 99237f8a4cc3
- 1mo ago First seen · 156 lines · 108 tokens per session scan A 515ad31b7816
eval-init is a skill published in the GitHub repository claude-code-expert/carve-harness (15 stars, last pushed 1mo ago), licensed MIT. It adds 108 tokens to every session and 3,524 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.
skill-benchmark
Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.