Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/Dannykkh/skill-olympusnpx agentmods add skills/dannykkh/skill-olympus/zeusWrote 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/dannykkh/skill-olympus/zeus)<a href="https://agentmods.dev/skills/dannykkh/skill-olympus/zeus"><img src="https://agentmods.dev/badge/skills/dannykkh/skill-olympus/zeus/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/dannykkh/skill-olympus/zeus"><img src="https://agentmods.dev/badge/skills/dannykkh/skill-olympus/zeus.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 7 findings, 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 265 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.
- medium Agent Snooping · line 432 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.
- medium Agent Snooping · line 265 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.
- medium Agent Snooping · line 432 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.
- medium MCP Rug Pull · line 355 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium Agent Snooping · line 431 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.
- medium Agent Snooping · line 436 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 | $0.00072 | $0.07657 |
| Opus 5.5 | $0.00029 | $0.03063 |
| Sonnet 5.5 | $0.00014 | $0.01531 |
| Haiku 4.5 | $0.00007 | $0.00766 |
Grade A, and why
zeus 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 16d 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 — 445 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Zeus (제우스) — Zero-Interaction Full Pipeline
"이런 프로그램 만들어줘" 한 줄이면 설계부터 구현, 테스트까지 전부 자동으로.
Pipeline Overview
/zeus "쇼핑몰 만들어줘. React+Spring Boot"
│
▼
Phase 0: Description Parsing ─── 산업군, 기술스택, 기능 추출
│
▼
Phase 1: Planning (zephermine) ─ 합성 인터뷰 → 26단계 설계 자동 완료
│
▼
Phase 2: Implementation (agent-team) ─ 포세이돈 PM이 섹션 기반 병렬 구현
│
▼
Phase 3: Verification (argos) ─ 설계 대비 준공검사
│
▼
Phase 4: Docker Setup (docker-deploy) ─ Docker 환경 구성 + 포트 충돌 해결
│
▼
Phase 5: Testing (minos) ─ E2E 테스트 + Healer 루프
│
▼
Phase 6: Final Report ─────────── docs/zeus/zeus-report.md 생성
CRITICAL RULES
- 메타 요청 구분 — 사용자가 Zeus 자체를 점검/수정/설명하라고 한 경우에는 제품 파이프라인을 시작하지 말고 이 SKILL.md와 관련 런타임만 점검한다. 제품/기능을 "제우스로 진행"하라는 요청일 때만 Phase 0~6을 실행한다.
- NEVER call interactive question tools — 모든 결정은 자동선택 규칙으로 처리 (
Recommended우선, 없으면 자동 응답 테이블) - 절대 멈추지 않는다 — 에러 시 기록하고 계속 진행
- Phase 완료 즉시 다음 Phase 시작 — "다음을 진행합니다" 같은 중간 보고 금지. Phase 0→1→2→3→4→5→6을 한 턴에 연속 실행
- [ZEUS-AUTO] 태그 + 결정 장부(Decision Ledger) — 질문을 사전에 하지 않는 대신 결재를 사후로 미룬다.
자동 결정에는 태그와 분류를 표시:
[ZEUS-AUTO:mechanical]— 정답이 하나인 결정 (기본값 선택, 폴백 등). 기록만.[ZEUS-AUTO:taste]— 합리적인 사람이 다르게 판단할 수 있는 결정 (디자인 톤, 기능 우선순위 등). 결정 장부 형식으로 기록: 결정 + 근거 + 기각한 대안 + 되돌리는 법. Phase 6 리포트의 Decision Ledger 섹션에 모아서 표시 — 사용자가 사후 결재(검토 후 뒤집기)할 수 있게 유지. 가능하면 되돌리기 쉬운(reversible) 기본값을 우선 선택한다.- 예:
[ZEUS-AUTO:taste] 인증 JWT 선택 | 근거: SPA 표준, 무상태 확장 | 대안: 세션(서버 상태 필요) | 되돌리기: auth 모듈 교체, API 계약 변경 없음
- 재개 지원 — docs/zeus/zeus-state.json으로 중단 지점부터 재개
- 모든 Phase 강제 실행 — Phase 0~6 모두 최소 1회 실행 시도 필수. "건너뜀"은 물리적 불가(Docker 미설치 등)일 때만 허용하며, 그 경우에도 폴백 경로를 실행
- 완료 전 최종 응답 금지 — Phase 0~5 실행 증거가 없으면 "다음 구현 순서"를 사용자에게 넘기지 않는다. 실행 가능한 다음 작업은 즉시 수행하고, 불가능하면
zeus-state.json과 핸드오프를 남긴다. - 스코프 게이트 (pass/reduce/hold) — zero-interaction에는 스코프 확장을 승인해줄 사용자가 없으므로,
plan.md/sections/확정 이후 계획에 없던 태스크·기능·신규 의존성을 추가하기 전에 자문 3개를 통과시킨다:- 사용자의 한 줄 목표에 직접 기여하는가? / 지금 필요하게 만든 관찰된 근거(실패한 테스트, 감리 지적, 명시 요청)가 있는가? / 기존 산출물 재사용이나 더 작은 변경으로 같은 결과가 나오는가?
- 판정은 셋 중 하나로 결정 장부에 기록:
pass(최소 범위로 진행) /reduce(축소 후 진행) /hold(실행 보류 — Phase 6 리포트 Decision Ledger에 Deferred로 표시, 사용자가 사후 승격) - 과적용 금지: 계획된 섹션 구현·버그 수정·Phase 폴백 실행은 게이트 대상이 아니다 (폴백은 스코프 축소이지 확장이 아님). 게이트가 규칙 6(모든 Phase 강제 실행)을 약화시키지 않는다.
- 예:
[ZEUS-AUTO:taste] 스코프 게이트 hold: 관리자 통계 차트 | 근거: 한 줄 요구·섹션에 없음, 감리 지적 아님 | 되돌리기: Deferred에서 승격 후 재실행
What ships with it
9 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.
- agents/openai.yaml 267 B
- commands/zeus.md 5.2 KB
- references/archive-procedure.md 1.6 KB
- references/auto-interview-generator.md 6.7 KB
- references/autopilot-defaults.md 8.1 KB
- references/description-parser.md 5.8 KB
- references/docker-setup.md 4.3 KB
- references/final-report-format.md 3.6 KB
- references/phase-transitions.md 7.5 KB
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.
- 16d ago Changed 87f90798d0c5
- 1mo ago First seen · 445 lines · 72 tokens per session scan A c6fe47a1c442
zeus is a skill published in the GitHub repository Dannykkh/skill-olympus (7 stars, last pushed today), licensed MIT. It adds 72 tokens to every session and 7,657 once invoked, about $0.0003 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 2026-09-06.
Other skills, from other repositories
factory-testing
Testing conventions across builds. Vitest + Playwright as the locked stack; tests co-located in tests next to the code they verify; shared test primitives (provider wrappers, mock factories) centralized in src/lib/test-utils.tsx; coverage thresholds enforced at the merge gate (CI ownership lives in factory-ci.md).…
audit-runner
Autonomous audit runner (separate sessions, no context limit). Use when: "audit-runner", "autonomous audit", "audit automatically".
refactor-team
Launch parallel refactoring agents: one restructures code in an isolated worktree, one writes tests to lock behavior, one reviews the result. Use when: "refactor team", "refactor-team", "safe refactor", "restructure code".
Web: Walkthrough
The sign-off walkthrough: discover every open state, plan every route on every intended device, capture through asset-shot under the capture naming rule, gate the results and record what failed.
debug-team
Launch parallel debugging agents for systematic root-cause analysis. Use when: "debug team", "help me debug", "debug-team", "find the bug".
e2e-testing
Guide for running end-to-end tests of the Qwen Code CLI, including headless mode, MCP server testing, and API traffic inspection. Use this skill whenever you need to verify CLI behavior with real model calls, reproduce user-reported bugs end-to-end, test MCP tool integrations, or inspect raw API request/response…