Claude DevfleetSAFE
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Overview
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
e9ca581a4f44OBSERVED · 2026-09-19Install
Commands as the repository documents them. They are shown, not run.
claude mcp add devfleet --transport http http://localhost:18801/mcp
Host compatibility
What the documentation claims. We have not run a compatibility test.
| Host | Status | Notes |
|---|---|---|
| claude-code | mentioned |
What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
--- name: claude-devfleet description: Claude DevFleet経由でマルチエージェントコーディングタスクをオーケストレーション — プロジェクトを計画し、分離された作業ツリー内で平行エージェントを派遣し、進捗を監視し、構造化レポートを読む。 origin: community --- # Claude DevFleet マルチエージェント オーケストレーション ## 使用時期 このスキルは、複数のClaude Codeエージェントをコーディングタスクで並行して作業するように派遣する必要があるときに使用します。各エージェントは完全なツール機能を備えた分離されたgit作業ツリーで実行されます。 実行中のClaude DevFleetインスタンスが必要で、MCP経由で接続: ```bash claude mcp add devfleet --transport http http://localhost:18801/mcp ``` ## 動作方法 ``` User → 「認証とテスト付きのREST APIを構築」 ↓ plan_project(prompt) → project_id + mission DAG ↓ 計画をユーザーに表示 → 承認を取得 ↓ dispatch_mission(M1) → エージェント1は作業ツリーで生成 ↓ M1完了 → 自動マージ → M2を自動派遣(M1に依存) ↓ M2完了 → 自動マージ ↓ get_report(M2) → files_changed、what_done、errors、next_steps ↓ ユーザーに報告する ``` ### ツール | Tool | Purpose | |------|---------| | `plan_project(prompt)` | AIが説明をミッションチェーン付きプロジェクトに分割 | | `create_project(name, path?, description?)` | プロジェクトを手動で作成、`project_id`を返す | | `create_mission(project_id, title, prompt, depends_on?, auto_dispatch?)` | ミッションを追加。`depends_on`はミッションIDの文字列のリスト。`auto_dispatch=true`で依存関係が満たされたとき自動開始。 | | `dispatch_mission(mission_id, model?, max_turns?)` | ミッション上でエージェントを開始 | | `cancel_mission(mission_id)` | 実行中のエージェントを停止 | | `wait_for_mission(mission_id, timeout_seconds?)` | ミッション完了までブロック | | `get_mission_status(mission_id)` | ブロックなしでミッション進捗をチェック | | `get_report(mission_id)` | 構造化レポートを読む | | `get_dashboard()` | システム概要:実行中のエージェント、統計 | | `list_projects()` | すべてのプロジェクトをブラウザ | | `list_missions(project_id, status?)` | プロジェクト内のミッションをリスト | ## ワークフロー 1. **計画**:`plan_project(prompt="...")`を呼び出す → `project_id` + ミッションリスト 2. **表示**:ユーザーにミッション計画を表示 3. **派遣**:最初のミッションで`dispatch_mission()`を呼び出す 4. **監視**:`get_mission_status()`で進捗をチェック 5. **報告**:`get_report()`で完了時の報告 ## 例 ### フル自動:計画と起動 1. `plan_project(prompt="...")` 2. 最初のミッションをDispatch 3. 残りのミッションは依存関係に基づいて自動Dispatch 4. 完了したらユーザーに報告
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | NA |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (0)
No findings outside the package's declared scope.
Gates applied: no_behavioural_pass.
e9ca581a4f44full audit observations/trust-audit/skill/affaan-m__claude-devfleet.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-09-19 | e9ca581a4f44 | SAFE | B | 89 | first audit |
Questions
What does the Claude Devfleet skill do?
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Is Claude Devfleet safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Claude Devfleet access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does Claude Devfleet work with?
Its documentation mentions claude-code. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (e9ca581a4f44), read on 2026-09-19. The repository is watched, and a new audit runs when it changes — this is the first audit.