minimax-ai/openclaw-acp-bridge
Bridge MiniMax Code to OpenClaw-mcode-ACP for true peer-to-peer collaboration. Use the inbox protocol to read messages, push progress, ask blocking questions, and answer peer questions, instead of one-shot master/slave task calls. The v0.3.0 release adds the goudan-side companion (acp-inbox-bridge Skill + scripts/acp_inbox.py wrapper + scripts/test_inbox_goudan.py smoke) so the OpenClaw main agent can also drive the inbox from the goudan perspective; the mavis-side Skills (acp-collab, acp-task-dispatch) are unchanged.
Use the OpenClaw-mcode-ACP inbox to collaborate peer-to-peer with goudan (OpenClaw main session) instead of one-shot task calls. Read incoming messages, push progress, ask blocking questions, and answer peer questions. Use when working on multi-step tasks that span MiniMax Code and OpenClaw sessions, when uncertain decisions need goudan's input, or when long-running work should report progress back to the parent session.
Use the OpenClaw-mcode-ACP inbox from inside an OpenClaw main session (goudan) to PROACTIVELY send messages, ask blocking questions, and answer mavis-side questions. Differs from acp-collab (mavis / MiniMax Code perspective) — this is the goudan-side companion that uses the same client/_acp_client.inbox_* HTTP transport and inherits the same loopback-only, no-redirect, token-via-env security model. Use when a multi-step task spans the OpenClaw main session and a MiniMax Code peer, when goudan should report progress or findings to mavis, when goudan needs to ask a clarifying question that requires a mavis-side tool (e.g. reading ~/.minimax/), or when mavis has asked a question and goudan must answer it before continuing.
Dispatch a self-contained task to the OpenClaw-mcode-ACP HTTP server from inside MiniMax Code. Use when a task should be persisted, retried, observed over time, or processed by a worker pool instead of the current MiniMax Code session.