The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources including YouTube channel descriptions, X/Twitter profiles, blog content, and podcast metadata.
Ingestion points: SKILL.md (Steps 3 through 7) defines the collection of influencer data from various third-party platforms.
Boundary markers: The instructions do not specify any delimiters or safety prompts to prevent the agent from following instructions embedded in the discovered content.
Capability inventory: The skill is restricted to web searching and markdown report generation. It does not possess file-writing, shell execution, or system modification capabilities.
Sanitization: There is no evidence of sanitization or filtering of the external content before it is processed for scoring and summarization.
[EXTERNAL_DOWNLOADS]: The skill requires active internet access to perform searches and fetch metadata from social platforms. While these are downloads of public data for analysis, they constitute a persistent connection to external services.

influencer-discovery