CawCut App Run

Entry point for anything CawCut-App related: discovery ("what can this app do", "what apps are available"), catalog browsing, and running Apps. Wraps cawcut app list, cawcut app describe, and cawcut app run. New media production is routed by cawcut-workflow; this skill owns selected-App inspection, inputs and execution.

Skill files are English-only. SKILL.md and references/ must not contain Chinese or other non-English prose (including YAML description trigger phrases). Match the user's language only in live replies via reply_language.

For media inputs run cawcut asset upload once and keep the returned asset_ids; reuse them instead of uploading again. Retain IDs and returned URLs from App/Product/stage results before later processing.

Step -1 — Handoff state (check first)

If cawcut-workflow already selected an App for this goal, carry its selection, catalog evidence, user inputs and asset IDs forward. Proceed to live describe and input collection without repeating the picker. Fetch current catalog data when resolving an ID or handling access changes, preserving the selected App. If App inspection proves unsuitable, return the reason and evidence to workflow planning; the canvas branch resumes without another catalog/picker loop.

Otherwise, continue at Step 0.