Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add TOKTOKHAN-DEV/agent-company --skill review-specgit clone --depth 1 https://github.com/TOKTOKHAN-DEV/agent-companyWrote 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/toktokhan-dev/agent-company/review-spec)<a href="https://agentmods.dev/skills/toktokhan-dev/agent-company/review-spec"><img src="https://agentmods.dev/badge/skills/toktokhan-dev/agent-company/review-spec/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/toktokhan-dev/agent-company/review-spec"><img src="https://agentmods.dev/badge/skills/toktokhan-dev/agent-company/review-spec.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00002 | $0.00569 |
| Opus 5 | $0.00001 | $0.00284 |
| Sonnet 5 | $0.00000 | $0.00114 |
| Haiku 4.5 | $0.00000 | $0.00057 |
Grade A, and why
review-spec 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 10d 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.
What it actually says
review-spec
명세는 한 번 쓰고 끝나지 않는다. 심사 기준이 갱신되고, 구현하면서 알게 된 것이 생긴다.
1. 대상을 정한다
ls specs/*.md
pnpm preflight # 어떤 명세가 error 인지
2. 심사 체크리스트와 대조한다
wiki/04-review-checklist.md 를 열고, 이 기능이 건드리는 영역을 위에서부터 훑는다.
특히 놓치기 쉬운 것:
| 영역 | 자주 빠지는 항목 |
|---|---|
| 내비게이션 | 뒤로가기 버튼이 두 개 동시에 보이지 않는가 |
| 사용성 | 진입하자마자 바텀시트가 뜨지는 않는가 |
| 사용성 | 다른 앱 설치나 자사 서비스로 유도하지 않는가 |
| 권한 | 권한이 거부돼도 기능이 도는가 |
| 로그인 | 토스 로그인 외 다른 로그인 수단을 만들지 않았는가 |
| 결제 | 취소하면 주문 화면으로 돌아오는가 |
| 광고 | 인트로·로딩·팝업 같은 임시 화면에 광고를 넣지 않았는가 |
3. 구현과 어긋난 곳을 찾는다
grep -n "path:" apps/miniapp/src/App.tsx # 실제 라우트
ls apps/miniapp/src/screens/ # 실제 화면
명세에 없는 화면이 있으면 둘 중 하나가 틀린 것이다. 화면을 지우거나 명세를 갱신한다. 임의로 판단하지 말고 사람에게 어느 쪽인지 묻는다.
4. 반려를 명세로 되돌린다
심사에서 반려됐다면 그 사유를 수용 기준으로 추가한다. 같은 이유로 두 번 반려되지 않게 하는 유일한 방법이다.
## 수용 기준
- [ ] (2026-08-07 반려 반영) 광고 시청 중 배경음이 멈춘다
완료 확인
pnpm preflight
출력
- 검토한 명세
- 추가한 수용 기준
- 명세와 구현이 어긋난 곳 (사람의 판단이 필요한 것)
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.
- 10d ago First seen · 66 lines · 2 tokens per session scan A 6aed0483c83c
review-spec is a skill published in the GitHub repository TOKTOKHAN-DEV/agent-company (77 stars, last pushed 14d ago), licensed MIT. It adds 2 tokens to every session and 569 once invoked, about $0.0000 per session on Opus 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
fix
Fix phase workflow. Fixes issues identified during code review, based on the triaged review output. Use after a review phase is complete and the human has triaged the findings, or when the user says "fix", "fix the issues", "address review feedback", or "apply fixes". Also use when the user says "fix these" after…
review
Code review workflow. Reviews the build output for quality, conventions, performance, security, and test coverage. Use after a build phase is complete, or when the user says "review", "check the code", "PR review", "quality check", or "is this good". Also use when the user wants a second opinion on code that was just…
review-loop
Run the adversarial verification loop — implement, then hand the change to a fresh checker that did not write it, fix what it finds, and re-dispatch until APPROVE. Use before claiming any behavioural change is done, and on requests like "review loop", "adversarial review", "independent review", "get this verified"…
security-review
Perform a focused security review of pending git changes to identify high-confidence security vulnerabilities with real exploitation potential. Use this skill when the user asks for a security review, security audit, vulnerability scan, or wants to check pending changes on a branch for security issues before merging.…
huggingface-llm-trainer
Train or fine-tune language models with TRL or Unsloth on Hugging Face Jobs, including SFT, DPO, GRPO, reward models, and GGUF conversion. Use for cloud LLM training; use huggingface-vision-trainer for vision tasks.
huggingface-vision-trainer
Train object-detection, image-classification, or SAM segmentation models on Hugging Face Jobs. Use for vision fine-tuning and evaluation; use huggingface-llm-trainer for language models.