Qianliu AI Testing

Problem

This skill targets test workflows on the Qianliu TP platform: checking test resources, polling task status, and collecting reports can be scattered across manual API calls. It groups testbed discovery, execution-host discovery, task execution, status tracking, and report retrieval into a consistent interface, which fits an agent-driven automation loop.

How It Works

The skill works through TP platform APIs. The main path is:
- Resource discovery: call GET /api/resource/testbeds and GET /api/resource/exec-hosts to inspect available testbeds and execution hosts.
- Task execution: configure task parameters in workspace/.qianliu/.qianliu-aitest/tp-aitest-config.yaml, then call POST /api/test/execute to start a test.
- Status tracking: query GET /api/test/status/{taskId} for execution state, then fetch the result with GET /api/test/report/{taskId}.

Before use, store the TP platform access token in ~/.qianliu/config.json. The token should be generated from TP platform API token settings and carry the required permissions. If a testbed is unavailable, an execution host cannot be reached, or the token is rejected, check resource status, network connectivity, and token expiry.