Claudio: run Claude Code as a service
说明与用途
Claudio 将 Claude Code CLI 转变为带有 REST API 和 Web UI 的服务,实现远程聊天、并行编排、多账户管理和安全访问。
问题
Running Claude Code from a terminal limits remote access, parallel execution, and tracking of multiple agents.
结果
Claudio transforms Claude Code into a service, enabling remote chat, parallel orchestration, multi-account management, and secure access, all while preserving the CLI’s surface.
工作流
- 1. Install Claudio (Docker, desktop app, or from source).
- 2. Sign in with Claude account via setup assistant.
- 3. Start a chat via UI or POST /api/chats with prompt and cwd.
- 4. Interact with chat, send messages, permissions, or stream via /api/chats/:id/stream.
- 5. For orchestrations, POST /api/orchestrations with a task graph; monitor via UI or events.
- 6. Manage multiple Claude accounts with CLAUDE_CONFIG_DIR per account; Claudio rotates usage.
- 7. Review changes using diff screen; use guards to enforce safety policies.
- 8. Secure the service by enabling authentication (token or OIDC) and placing a TLS proxy if exposed externally.
证据
Claudio puts the Claude Code CLI behind a REST API and a web UI.
A chat is one long-lived claude process fed over stdin, with --resume when the process is gone.
Workers run in parallel, each in its own git worktree, and their branches are merged into one integration branch.
Each Claude account gets its own CLAUDE_CONFIG_DIR and is signed in once. Claudio reads usage per window, moves new work to an account with quota before a limit is hit.
Claudio can add PreToolUse hooks that hold even in bypassPermissions: no push, no --no-verify, no recursive delete outside the worktree, no secret written.
- 记录日期: 2026-10-25T00:26:16.821Z