The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it instructs the agent to fetch and analyze content from untrusted external sources.
Ingestion points: The skill processes external data including creator posts, audience comments, and media kits accessed via tools such as social-fetch and watch-video, as well as manual 'mining' of community discussions.
Boundary markers: The instructions lack explicit delimiters or 'ignore embedded instructions' warnings for the data fetched from social media platforms.
Capability inventory: The skill has access to read project context files (e.g., .agents/product-marketing.md), execute network-enabled tools (social-fetch), and generate detailed marketing plans or outreach scripts.
Sanitization: While the skill includes detailed vetting procedures for 'fake-follower' detection and engagement quality, it does not specify technical sanitization or filtering to prevent the agent from being influenced by hidden instructions in the ingested content.

influencer-marketing