원클릭으로 Manus에서 모든 스킬 실행

autonomous-fleet-core

스타0

포크0

업데이트2026년 7월 8일 19:57

The portable, tool-agnostic ENGINE for running fully-autonomous multi-agent engineering jobs. Shipped mission skills (doc-sync, test-coverage, adversarial-review-and-fix) invoke THIS engine plus exactly one ADAPTER (orca, claude-code, grok, or another runtime). Eighteen additional missions are documented under docs/exploratory/missions/ and re-promote on real-run evidence. This core holds everything that does NOT depend on orchestration tool: self-orientation, fully-autonomous coordinator behaviour with file-ledger boolean gates, context-handoff to survive compaction, the worker-placement DECISION LOGIC (dependent vs independent), the PR-per-task pipeline with commits-preserved + conflict-aware merge + worktree cleanup, the empirical risk tiers, safety rails, secret hygiene, and commit/authorship policy. It speaks in PRIMITIVES; the ACTIVE ADAPTER maps each primitive to its tool's real commands. Load with a mission skill and one runtime adapter — do not run alone.

설치

Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.

Manus에서 실행

출처

ravidsrk

ravidsrk/autonomous-fleet

GitHub 저장소 열기 Creator 저장소 보기

다운로드

Manus에서 실행

Autonomous Fleet — Core Engine (tool-agnostic)

You are the COORDINATOR for an autonomous multi-agent run. You are a THIN LOOP-HOLDER: create tasks, spawn workers, dispatch, wait, answer worker questions from defaults, sequence the pipeline, decide what runs next and how parallel. You do NOT review, code, or merge yourself — all dispatched. Context stays light; the source of truth is the ledger FILE, not your memory.

Required composition

Do not run this skill alone. A mission skill defines the work; an adapter skill defines runtime mechanics.

Instructions

Read and follow references/engine.md in full before coordinating any run — it is the slim ALWAYS-READ core capped at 300 lines. Its trigger map tells you when to load references/engine-autonomy.md, references/engine-workers.md, references/engine-review.md, and references/engine-recovery.md (issue #84). It contains the engine identity, coordinator loop contract, ledger boolean gates, the 4-layer index, and trigger stubs for full doctrine.

For how missions, worker skills, optional skills, and campaigns compose, read references/composition.md, references/community-skills.md, references/fleet-outcome.md, and references/runtime-goals.md for native /goal / update_goal binding. Per-repo defaults: setup-autonomous-fleet → docs/agents/fleet-config.md. Mission chains and conditional DAGs use fleet-program — not a second mission loaded alongside the first.

For review missions emitting structured findings (adversarial-review-and-fix and any future reviewer phase), see references/review-findings.md: the JSON schema, the source-verification CLI, the fix-strategy/confidence gating rules. For RUNTIME enforcement of the engine's EVID/WT_CLEAN/e2e_verified disciplines via a Claude Code Stop hook, see references/strict-mode.md: opt-in install of the stop-verify gate that refuses session termination without verifiable evidence on disk.

Every run that emits first-class artifacts (findings, blind-fix files, verifier summaries) leaves a manifest-audited trail under .fleet/runs/<run_id>/. See references/run-archive.md for the manifest scheme and the archive_enabled gate (a status: done outcome is rejected if its archive doesn't validate).

Substrate distribution (skills-install mode). The Python enforcement substrate (the validators above plus lib/) ships WITH this skill under assets/substrate/ — on a skills-installed repo it lives at .agents/skills/autonomous-fleet-core/assets/substrate/ and each CLI runs standalone with python3 (deps: assets/substrate/requirements.txt; version pinned in assets/substrate/substrate-manifest.json). In a framework clone, prefer the canonical scripts/ copies. Shell wrappers (preflight.sh, validate-fleet-outcome.sh, run-sandboxed.sh) remain clone-only for now. setup-autonomous-fleet records the resolved SUBSTRATE_PATH in docs/agents/fleet-config.md.

Ports of Agent Orchestrator mechanisms (nudge dedup, stacked PR, hook-signal, review supersede) without adopting AO's daemon: references/ao-adoptions.md.

Primitives (summary)

The active adapter must implement: PLACE, SPAWN_WORKER, DISPATCH, WAIT, INSPECT, WORKER_DONE / ASK / REPLY, OPEN_PR / MERGE_PR / CLEANUP, SYNC_TASK_STATE. When the host supports goal mode, also implement: SET_GOAL, UPDATE_GOAL, GOAL_COMPLETE, GOAL_BLOCKED, and LOOP_POLL (host-native scheduler; Orca exempt — see runtime-goals.md). Optional primitive 14, CONTINUE_WORKER (re-attach an existing resumable session for an in-flight task), is implemented by adapters whose runtime exposes a restore command and ALIASED to SPAWN_WORKER (idempotent relaunch) otherwise — see engine-workers.md.

이 저장소의 다른 Skills

같은 저장소

devex-audit

ravidsrk/autonomous-fleet

[Tier 2 · moderate autonomy · full review gate · developer-experience] Run a live developer experience audit with frozen DX scorecard and ranked documentation gaps. Maps gstack /devex-review and /plan-devex-review into fleet ledgers with optional document-generate follow-ups. Use when onboarding friction, CLI help, or docs TTHW need evidence before ship. Trigger on: "DX audit", "developer experience review", "devex mission", "audit onboarding", "score the getting started flow", "plan devex review".

2026-07-050

landing-page-convergence

ravidsrk/autonomous-fleet

[Tier 2 · moderate autonomy · full review gate] Force a production landing page that has DIVERGED from an approved design into full fidelity with that design - section by section, with a named divergence checklist as the forcing function so it converges instead of drifting. Use when a live page no longer matches its design export and needs to be brought back into line (a single page/site, not a whole app - that's design-integration). Extracts the design system, diffs production against it, and closes each divergence as its own PR. Runs via the autonomous-fleet-core engine. Trigger on: "make the landing page match the design", "the production page diverged from the design", "bring the landing page into parity", "fix the landing page to match the mockup".

2026-07-050

legacy-rebuild

ravidsrk/autonomous-fleet

[Tier 3 · high blast radius · expect rework · run deliberately, review the floor + architecture] Adversarially review a legacy app, research the best current architecture, and rebuild it end to end on a modern foundation while PRESERVING everything it currently does. Use when an app uses outdated everything (old deps, poor architecture, e.g. JS-inlined-in-HTML, no build/module system) and needs a real modernization, not a patch. Incremental and shippable per PR — NOT a big-bang rewrite — rebuilt against a captured behaviour floor so nothing is silently lost. Runs via the autonomous-fleet-core engine. Trigger on: "rebuild this legacy app", "modernize the whole codebase", "this uses all old versions, rebuild it properly", "re-architect end to end".

2026-07-050

product-framing

ravidsrk/autonomous-fleet

[Tier 2 · moderate autonomy · full review gate · pre-build] Turn vague product intent into a frozen, review-approved product spec before any implementation. Runs gstack-style office-hours forcing questions plus CEO/design/eng plan reviews (via optional community skills), then emits docs/product-spec.md as the single source of truth for downstream build missions. Use when the user has an idea, wedge, or feature direction but no frozen spec yet. Trigger on: "frame this product", "office hours on this idea", "is this worth building", "freeze the product spec", "run plan reviews before we code", "product framing mission".

2026-07-050

release-document

ravidsrk/autonomous-fleet

[Tier 2 · moderate autonomy · full review gate · post-ship docs] Post-ship documentation sweep and deploy verification checklist after a version lands. Maps gstack /document-release, /ship, /land-and-deploy, and /canary into fleet frozen artifacts. Use after merge or deploy when changelog, user docs, and canary notes need closure. Trigger on: "document this release", "post-ship docs", "release documentation mission", "update docs after ship", "canary doc sweep".

2026-07-050

security-cso-audit

ravidsrk/autonomous-fleet

[Tier 2 · moderate autonomy · full review gate · security-specialist] Run a gstack-style CSO security audit (secrets archaeology, supply chain, CI/CD, OWASP, STRIDE, LLM/skill supply chain) with daily or comprehensive modes, freeze findings, then remediate confirmed issues one PR at a time with skeptic narrowing. Distinct from general adversarial-review-and-fix by infrastructure- first CSO lenses. Trigger on: "CSO audit", "security audit mission", "OWASP review", "threat model this repo", "run /cso as a fleet mission", "supply chain security pass".

2026-07-050

name	autonomous-fleet-core
description	The portable, tool-agnostic ENGINE for running fully-autonomous multi-agent engineering jobs. Shipped mission skills (doc-sync, test-coverage, adversarial-review-and-fix) invoke THIS engine plus exactly one ADAPTER (orca, claude-code, grok, or another runtime). Eighteen additional missions are documented under docs/exploratory/missions/ and re-promote on real-run evidence. This core holds everything that does NOT depend on orchestration tool: self-orientation, fully-autonomous coordinator behaviour with file-ledger boolean gates, context-handoff to survive compaction, the worker-placement DECISION LOGIC (dependent vs independent), the PR-per-task pipeline with commits-preserved + conflict-aware merge + worktree cleanup, the empirical risk tiers, safety rails, secret hygiene, and commit/authorship policy. It speaks in PRIMITIVES; the ACTIVE ADAPTER maps each primitive to its tool's real commands. Load with a mission skill and one runtime adapter — do not run alone.
license	MIT
compatibility	Requires git and gh CLI in the target repository
metadata	{"author":"ravidsrk","version":"1.3.4","fleet-component":"core"}

Autonomous Fleet — Core Engine (tool-agnostic)

Required composition

Do not run this skill alone. A mission skill defines the work; an adapter skill defines runtime mechanics.

Instructions

Ports of Agent Orchestrator mechanisms (nudge dedup, stacked PR, hook-signal, review supersede) without adopting AO's daemon: references/ao-adoptions.md.