Pudu Cloud API

Problem Addressed

When integrating Pudu robot Open APIs, engineers often need to manage credentials, cluster endpoints, path prefixes, parameter shapes, and response validation at the same time. If these details are scattered across docs, scripts, and logs, failures can surface as 401, 403, 404, or an HTTP success that is still a business failure.

How It Works

• Credential checks: Prefer PUDU_API_APP_KEY and PUDU_API_APP_SECRET; prompt for missing values and mask ApiAppSecret instead of echoing full secrets.
• Cluster confirmation: Map PUDU_API_CLUSTER values such as cn, sea, de, and us to the corresponding hostname; require explicit user selection when the cluster is ambiguous instead of defaulting to cn.
• Endpoint lookup: Read the matching assets/*.json spec files for robot status, delivery, errands, lifting and tray tasks, cruise guidance, calls, Flash Cabinet, media, control reporting, and analytics.
• Request execution: Prefer pudu-request.js, with fallbacks to Python, Go, Java, and C# scripts when needed; include the required /pudu-entry or /data-board path prefix defined by the spec.
• Result rendering: Treat the call as successful only when the HTTP status is 200 and message === "SUCCESS"; otherwise report the error code and remediation hint.

Boundaries

This skill is intended for developers who already have valid Pudu Open Platform credentials and want to call Open APIs directly. It does not replace API permission provisioning, and parameters should not be guessed from natural language alone. Confirm sn, shopId, cluster, and API version against the spec files.