The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it is designed to ingest and process external content from app store metadata and local context files.
Ingestion points: The skill reads app-ads-context.md and utilizes the get_app and get_app_screenshots tools to retrieve external application information.
Boundary markers: There are no explicit boundary markers or instructions provided to the agent to disregard potential malicious directives embedded within the external app data or screenshots.
Capability inventory: The skill includes tools to list advertisement identities (tiktok_ads_list_identities) and provides strategies for managing Spark Ad batches.
Sanitization: The instructions do not define sanitization or validation logic for data retrieved from external sources before it is processed by the agent.

tiktok-creative-strategy