The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, specifically mobile app store listings, to generate creative assets and marketing recommendations. This introduces a surface for indirect prompt injection if a listing contains adversarial content intended to influence the agent's output. * Ingestion points: The skill ingests untrusted content from the app_url parameter in generate_app_ad_creative and store data in meta_app_ad_intelligence. * Boundary markers: The instructions do not define specific delimiters for processed external text but instruct the agent to perform analysis before generation. * Capability inventory: The skill possesses the capability to perform network operations and modify ad campaign settings through specialized tools (meta_ads_create_campaign, tiktok_ads_create_adgroup, etc.). * Sanitization: The risk is mitigated by explicit instructions requiring the agent to obtain human confirmation before executing any write operations or spend-changing actions (e.g., 'Always confirm with user before executing write operations').

app-growth