orchestrate · bloodfeast — скилл для Claude Code: координация агентов
Скилл orchestrate для Claude Code делит крупную задачу на направления и передаёт их субагентам. Он включается, когда работу нужно делегировать или контекста одного агента не хватает.
Apache-2.0 ★ 135 обновлён 2026-11-17
/plugin marketplace add bloodfeast/atlas
/plugin install atlas@atlasЧто делает orchestrate
orchestrate помогает главной сессии распределять работу между субагентами, проверять их результаты и объединять изменения. Для каждого направления агент ведёт журнал и передаёт следующему исполнителю handoff; новый агент продолжает с сохранённого состояния через recall.
Скилл учитывает лимиты контекста и ограничения числа агентов по типам. Atlas может поставить запуск в очередь, если свободного слота нет. После достижения лимита агент передаёт работу, а продолжение начинается с нового агента. Перед объединением результатов нужно самостоятельно запустить проверку или прочитать её вывод.
Какие задачи решает
- разбить крупную задачу на направления для субагентов
- передать работу новому агенту через handoff
- распределить запуски с учётом лимитов по типам
- проверить результат субагента до объединения изменений
Примеры запросов агенту
- Разбей эту задачу на направления и делегируй работу субагентам.
- Продолжи задачу с handoff и проверь результат перед объединением.
- Atlas сообщил о лимите агентов: распредели оставшуюся работу по очереди.
Как установить orchestrate
1Через маркетплейс плагинов
/plugin marketplace add bloodfeast/atlas/plugin install atlas@atlasКоманда ставит плагин atlas целиком, со всеми его скиллами
2Через Skills CLI (Vercel)
npx skills add bloodfeast/atlas --skill orchestrate -a claude-code3Вручную: положить файлы
~/.claude/skills/orchestrate/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/orchestrate/
1Через Skills CLI (Vercel)
npx skills add bloodfeast/atlas --skill orchestrate -a codex2Вручную: положить файлы
.agents/skills/orchestrate/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add bloodfeast/atlas --skill orchestrate -a cursor2Вручную: положить файлы
.cursor/skills/orchestrate/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install bloodfeast/atlas skills/orchestrate2Вручную: положить файлы
.github/skills/orchestrate/SKILL.md1Через Skills CLI (Vercel)
npx skills add bloodfeast/atlas --skill orchestrate -a gemini-cli2Командой платформы
gemini skills install https://github.com/bloodfeast/atlas --path skills/orchestrateВручную: положить файлы
.devin/skills/orchestrate/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Claude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы об orchestrate
Как установить orchestrate в Claude Code?
Выполни команды по порядку: /plugin marketplace add bloodfeast/atlas, затем /plugin install atlas@atlas.
Что делает orchestrate?
Скилл делит крупную работу на направления, передаёт их субагентам и помогает сохранить состояние между агентами. Перед объединением он требует проверить результат.
Что нужно для работы orchestrate?
Скилл использует atlas для журналов, handoff, восстановления состояния и управления лимитами агентов.
Исходный текст
SKILL.md · Apache-2.0
--- name: orchestrate description: Run a large piece of work as an orchestrator - split it into lanes, delegate each lane to a subagent with a lean brief, keep within atlas's per-kind agent caps and context limits, verify before merging, and start a fresh agent from the handoff whenever one reaches its limit. Use when work is bigger than one agent's context, when the user asks to delegate, orchestrate, parallelise or "have agents do it", or when an atlas agent-cap or context-limit message appears. --- # Orchestrating with atlas The main session plans, delegates, verifies and merges. Subagents do the work. atlas holds the state between them: lane journals, handoffs and recall. No agent has to carry another agent's context. ## Lanes - A **lane** is one stream of work with an id, for example `PARSER-2`. Each agent works in exactly one lane, and journals under it with the atlas `journal` tool: its brief, its decisions, its measurements, its commits, any questions, and what is next. - An agent that stops writes a **handoff** for its lane with the atlas `handoff` tool. The next agent starts from `recall lane=<id>`, so nothing is lost when an agent ends. - **Never resume an agent past its handoff.** Start a fresh one from the handoff. A short reply to an agent that is mid-task and below its handoff point is fine. - **Declare a lane** before its first launch: a `brief` journal entry in the lane with Kind, Lane, Priority and Uses lines, the same lines the launch prompt carries (see the brief, step 2). For example: ```text Kind: builder Lane: FEATURE-1 Priority: normal Uses: compile, worktree:feature-1 ``` Kind stays a property of each launch: the same lane later runs reviewers with `Kind: reviewer` and `Uses: read-only`. A builder launched with no `Lane:` line is let through, and atlas reminds you to declare its lane. ## Kinds and caps Every agent has a kind: - **builder** writes and compiles code; - **reviewer** reads and reviews, and does not build; - **research** researches or designs, and does not build. With agent caps on (`atlas settings orchestrate` asks for them), atlas denies a launch past its kind's cap. Setup derives the caps from the machine unless you answer them: compile slots of one per 16 hardware threads (at least 1, at most 4) and 1 gpu slot, then builders equal to the compile slots, as many reviewers, and half as many research agents (at least 1). A configured value always wins. When a launch is denied, do not retry it: atlas queues it (the reason names its key), and a task-notification lists what waits when an agent finishes. Then launch the next one whose kind has a free slot, with the same description and brief (`recall KEY` reads it); that removes it from the queue. `atlas queue` lists the queue, and `atlas queue drop KEY` removes a launch you no longer want. The kind comes from a `Kind:` line in the first 40 lines of the Agent launch prompt itself; a brief file the prompt names is not read. Without one, Explore and Plan agents count as research, an agent in its own worktree counts as a builder, and anything else gets the configured default kind (builder). So write the line. ## The brief Keep it lean. A brief is a work order, not a copy of the rulebook. 1. **Who and what:** the lane id, the goal in one or two sentences, and the deliverable (a branch, a PR, or a report). 2. **`Kind: builder`** (or `reviewer`, or `research`), and the work queue's lines: `Lane: <id>`, `Priority: urgent`, `normal` or `background`, and `Uses:` with the exclusive resources it needs (`gpu`, `compile`, `worktree:<name>`, one port or database such as `sqlport 15126`) or `read-only` for an agent that shares its lane. Never list repositories: each agent works in its own worktree. Only `gpu`, `compile`, `worktree:*` and the resources the configuration lists can hold a launch back; any other word (`blender`, unless it is configured) only adds a note when a running agent uses it too. atlas schedules on declared lines; it only warns on what it infers. A planned item's brief adds `Queue: <key>` so its launch takes it off the queue. 3. **The rulings that apply,** named by their recall key, a heading or a line hint, never pasted whole. "Read the whole rulebook" costs tens of thousands of tokens before any work starts. The agent pulls exactly what it needs with `recall`, `lookup` or `search`. Write the brief as compact Gherkin: a Context line, Given lines with the facts, a Scenario per acceptance criterion (its Then lines each a test that fails first), a Spike per research question (its Then names the answer: a table, a verdict against a bar, a number with its unit), one Rules line. Declare each key as a tag, in full every time (`@journal:LANE:0012`, pinned `@journal:LANE:0012#a1b2c3d4`), above Feature or above the scenario it serves; the agent's start lists each by key and title. Not `0012, 0013` after a key, nor `0140 onward`. 4. **Scope:** the files or modules it owns, what it must not touch, and what it is safe to do without asking. 5. **Verification:** the exact command that proves the work, and the numbers it must match. 6. **Journal and limits:** "Journal to lane `<id>`. From your handoff point, wrap up at the next clean stopping point and write your handoff; you must hand off by your hard limit." 7. **The report:** hand back the digest form. Lead with the outcome (PR, branch, merge, verdict) in the first lines, then the numbers, every open question under a `## Questions` heading, and every risk (a denied action, an unverified claim, a failure left open). File anything longer (the full notes, tables, logs) as a journal entry and cite its key. The report stays in the orchestrator's context for every later turn, so a line that changes nothing it does is a cost. A report over 28,000 bytes is sent back once: atlas files it whole and asks for the digest form. ## Context limits atlas reads each agent's real context size from its transcript and applies its kind's limits. The defaults on a 1M window are a handoff point at 285k and a hard limit at 450k, and both scale with the window. - **Past the handoff point,** the agent is told to wrap up: finish the step, commit, push, and write the handoff. - **Past the hard limit** (in enforced mode), every tool except `handoff` is denied, and so is a SendMessage to that agent. When such an agent finishes, atlas tells you to start a fresh agent from its handoff. Depending on the `fresh` setting, it tells you to start it or to ask the user first. The fresh agent's brief names the lane and says to start from `recall lane=<id>`. The main session follows the same rule. Near your own limit, journal and hand off, and offer the user a compaction point. ## Verify, then merge - **Do not take a report's word for a result.** Run the verification yourself, or read its output: test counts, eval lines, the diff. - **Review independently when it matters.** Give the reviewer the code and the claim, not your conclusion. Name reviewers by job (simplicity, cost), never by person. - **File each review** as a `journal` entry `review` with meta `role:`, `pr:`, `head:` (the sha it read), `verdict:` and `report:`. With `orchestrate.reviews` set, `gh pr merge` waits for every role's review of the PR's current head. - **Merge only what passed.** Then record the merge commit in the lane journal. ## Finish cleanly When a lane completes: - remove its worktree and its build target; - close the lane with its handoff; - say what moved, which questions wait on the user, and what is next. Ask the user design questions as multiple choice, with the recommended option first.
© Brandon White · Apache-2.0 · bloodfeast/atlas
Похожие скиллы
Все скиллы по теме «Настройка агента: память, контекст, субагенты» →Сначала — orchestrate от других авторов.