huawei-cloud-cloudrobo-dispatch

Windows / PowerShell: Examples use bash syntax. To run on Windows PowerShell:

• Flatten \ line continuations to a single line, or end lines with a backtick.
• Set env vars with $env:NAME="value" instead of export NAME="value".
• Single-quoted JSON '{"a":"b"}' works as-is.

Overview

The cloudrobo-dispatch skill manages embodied task dispatch via the robo-dispatcher service. It lets the agent create a task that runs on a robot (identified by robot_id) under an execution model (exec_model_id) inside a session (session_id), monitor it, cancel it, and retrieve the natural-language task result plus log items.

Applicable scenarios:

• Task dispatch — Create a task on a robot given a natural-language task description; always provide a stop condition (exec_constraints: default max_run_time=10 min, max_iter_num=100 steps)
• Task monitoring — List/show tasks in a session, filter by status/robot/infer-service
• Wait for task completion — Block until a dispatched task leaves RUNNING without manual polling (wait-task, the preferred way to wait for a task to finish)
• Task cancellation — Cancel a running/pending task
• Result retrieval — Get the task result (task + log_items) of a finished task
• Cross-skill orchestration — Combine robots (cloudrobo-robot) and inference models (cloudrobo-infer / cloudrobo-asset) into concrete robot-executable tasks.

Architecture:

Agent / LLM
    │
    ├── CLI  →  cloudrobo dispatch <command>
    └── SDK  →  DispatchClient (Python) — domain RoboDispatcherTaskManagement
                    │
                    ▼
              cloudrobo-service (REST API)
              /v1/robo-dispatcher/sessions/{session_id}/tasks*

All operations target the cloudrobo-service backend, are scoped to a session_id, and require a valid workspace_id. A session is a stable execution context; multiple tasks can be created within one session across time.

session_id: In the current version, session_id is identical to the current workspace's workspace_id. You can obtain it by reading the active workspace (cloudrobo workspace current) or the configured workspace_id, and reuse it directly as the --session-id. If a caller explicitly provides a different session_id, prefer confirming it against the current workspace.