social-content
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes a methodology for 'Reverse Engineering Viral Content' which involves scraping and analyzing large datasets (500-1000+ posts) from social media platforms. This process creates a surface for indirect prompt injection where malicious instructions embedded in the external content could influence the agent's behavior during analysis.
- Ingestion points: External social media post content scraped from LinkedIn, Twitter/X, and other platforms (SKILL.md).
- Boundary markers: Absent; the instructions do not define delimiters or protective instructions for the agent to distinguish external data from its core tasks.
- Capability inventory: The skill itself is documentation-based and contains no scripts, but the described workflow relies on the agent's core capabilities to ingest and analyze large volumes of text.
- Sanitization: Absent; no guidance is provided on filtering, validating, or sanitizing the data gathered from social platforms before processing.
Audit Metadata