The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using npm exec to interact with the @volcengine/kickart-open-mcp CLI tool for authentication, material uploading, and task submission. These operations are restricted to a specific vendor-owned package.
[EXTERNAL_DOWNLOADS]: The skill downloads the required CLI package from the official npm registry (registry.npmjs.org). As this is a well-known service and the package is provided by the skill author (Volcengine), this is considered standard operational behavior.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data including product URLs, product IDs, and user-provided creative prompts.
Ingestion points: User-provided inputs are collected in references/一键成片指南.md and references/视频参考指南.md.
Boundary markers: The skill mandates a two-step confirmation process in references/一键成片指南.md (Phase 0, steps 11 and 12), requiring the agent to display all collected information and obtain explicit user consent before proceeding.
Capability inventory: The skill has the ability to execute shell commands and read local media files via the KickArt CLI.
Sanitization: The skill performs local validation on file formats, sizes, and metadata before transmission.

byted-kickart-marketing-material-generator