agent-team

agent-team is a skill for Claude Code, Codex from Dannykkh/skill-olympus. It costs 132 tokens per session (9,571 once invoked), scanned A, original, MIT.

A guide for coordinating a team of coding agents working on one task. It covers when to split work into parallel assignments, how a lead agent coordinates them, and how to record shared knowledge.

In plain words
What is it for?
Use it to plan multi-agent tasks, assign independent work, coordinate results, and externalize decisions and findings.
Why use it?
It helps avoid duplicated work and prevents important decisions from remaining only in an agent’s memory. Parallel work is used when it can reduce the total effort.

Skill for Claude CodeCodex

Written for Claude Code and Codex: $ARGUMENTS substitution, but also agents/openai.yaml present. Also seen: reads .claude/ paths; mentions subagents; mentions Claude Code.

Part of the skill-olympus plugin — 101 skills, 7 commands, 5 MCP servers shipped together

Good fit Use it to plan multi-agent tasks, assign independent work, coordinate results, and externalize decisions and findings.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/dannykkh/skill-olympus/agent-team
Install

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.

Any agent
npx skills add Dannykkh/skill-olympus --skill agent-team
Clone the repo
git clone --depth 1 https://github.com/Dannykkh/skill-olympus

Made for: Claude Code, Codex.

Or install skill-olympus, the plugin that ships this one along with the rest of its 101 skills, 7 commands, 5 MCP servers.

Wrote 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.

agentmods badge for agent-team

README.md
[![agentmods](https://agentmods.dev/badge/skills/dannykkh/skill-olympus/agent-team/github.svg)](https://agentmods.dev/skills/dannykkh/skill-olympus/agent-team)
Your own site
<a href="https://agentmods.dev/skills/dannykkh/skill-olympus/agent-team"><img src="https://agentmods.dev/badge/skills/dannykkh/skill-olympus/agent-team/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.

agentmods 80×15 button for agent-team

Your own site · 80×15
<a href="https://agentmods.dev/skills/dannykkh/skill-olympus/agent-team"><img src="https://agentmods.dev/badge/skills/dannykkh/skill-olympus/agent-team.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 132 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 9,571 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00132 $0.09571
Opus 5.5 $0.00053 $0.03828
Sonnet 5.5 $0.00026 $0.01914
Haiku 4.5 $0.00013 $0.00957

Measured 3d ago against content hash 146213bb9a8a, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

agent-team 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 3d 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.

skills/agent-team/SKILL.md · 571 lines

How it starts

The opening of the file, as written. The whole thing — 571 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Team — Zephermine 섹션 병렬 실행

포세이돈(Poseidon): 젭마인 산출물을 받아 체계적으로 구현합니다. 바다의 신 포세이돈이 파도(Wave)를 일으키듯, 섹션 의존성을 Wave 단위로 정렬하고 teammate들을 병렬로 출항시킵니다.

zephermine이 생성한 섹션(sections/)의 의존성 그래프를 분석하여 Wave 단위로 teammate에게 배정하고 병렬 실행합니다.

다이달로스 vs 포세이돈

상황 사용할 도구
젭마인 없이 바로 구현 시작 다이달로스 (/daedalus) — 직접 리서치 → 제안 → 구현
젭마인 산출물(sections/) 기반 구현 포세이돈 (/agent-team 또는 /poseidon) — 섹션 파싱 → Wave → 구현

Lead(PM) 핵심 원칙

다이달로스의 PM 철학을 포세이돈 Lead에도 적용합니다.

1. 조건부 외주화 — 병렬 이득이 있을 때 Lead는 조율한다

Lead의 기억 공간이 전체 작전을 기억하는 유일한 곳이다. 코드까지 짜면 기억이 순식간에 꽉 찬다. 작업자가 가용하고 파일 범위가 독립적이면 코딩·리서치를 위임합니다. 작업자가 없거나 같은 파일을 순차 수정해야 하면 Lead가 실행자 역할도 소유합니다.

2. 기억 외부화 — 기억력을 믿지 마라

대화가 길어지면 오래된 내용이 자동 압축된다. 중요한 결정이 나올 때마다 activity log에 즉시 기록한다.

3. 체크리스트 완수 — 모든 Acceptance Criteria가 통과할 때까지 끝이 아니다

젭마인 산출물에는 섹션별 Acceptance Criteria(체크리스트)와 flow-diagrams(공정 도면)이 있다. teammate가 "완료"라고 보고해도 Lead가 직접 체크리스트를 대조하여 모든 항목이 통과할 때까지 반복한다. 완료 기준은 젭마인이 ID를 붙여 고정한 checklist.md(약속)이고, 이행은 Lead가 checklist-status.md 장부에 증거와 함께 기록한다. 형식·상태·증거 규칙의 정본은 젭마인 references/acceptance-checklist-guide.md이다. 대화로만 대조하면 컨텍스트가 압축될 때 함께 사라지므로 장부에 남긴다. 한 번 구현하고 끝내는 것은 PM이 아니라 실행자다.

Lead 운영 규율

Lead가 직접 하는 것:

  • 젭마인 산출물 검토 (plan, sections, flow-diagrams, acceptance criteria)
  • teammate 보고 수신 및 체크리스트 대조
  • 의사결정 + activity log 기록
  • teammate 배정/교체
  • 미통과 항목 → teammate에게 재지시
  • 네이티브 위임이 없거나 순차 실행이 더 안전한 작업의 직접 수행

작업자가 가용한 병렬 경로에서 Lead가 하지 않는 것:

  • 작업자와 같은 파일을 동시에 수정
  • 독립 작업자의 조사·테스트를 중복 실행
  • 외부 검증 없이 완료 선언

자기검증 3질문 — Wave 완료 보고 시 반드시 자문:

  1. 가장 어려운 결정이 뭐였나?
  2. Acceptance Criteria 중 위험한 항목은?
  3. 도면과 실제 구현이 일치하는가?

팀원 관리 원칙

규칙 설명
파일 영역 분리 같은 파일을 두 teammate가 동시에 수정 금지
idle 방치 teammate idle 알림이 와도 task 진행 중이면 절대 개입 안 함
교체 정책 다음 Wave가 이전 작업과 무관하면 → 새 teammate. 연장선이면 유지
이름 규칙 교체 시 같은 이름 재사용 불가

CLI별 실행 모드

native-first 원칙 (learned/020): 실행 엔진은 각 CLI의 네이티브 멀티에이전트에 위임하고, orchestrator MCP는 네이티브가 없는 구버전 CLI의 폴백 + hard file-lock 정책 레이어로만 남긴다.

Read the full file on GitHub · 571 lines

Files

What ships with it

8 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.

Changes

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.

  1. 3d ago Changed · +10 lines 146213bb9a8a
  2. 4d ago Changed · +14 lines 5c572f7ec51b
  3. 22d ago Changed · +8 lines 82b6b5ba86b9
  4. 1mo ago Changed · +17 lines · +45 tokens per session 30002f859555
  5. 1mo ago First seen · 522 lines · 87 tokens per session scan A 614fc27d9636

Subscribe to this mod's changes

agent-team is a skill published in the GitHub repository Dannykkh/skill-olympus (7 stars, last pushed 3d ago), licensed MIT. It adds 132 tokens to every session and 9,571 once invoked, about $0.0005 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 2026-08-31.