social-content
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill instructs the agent to analyze 500-1000+ scraped social media posts to identify viral patterns, which creates a potential surface for indirect prompt injection.\n
- Ingestion points: Untrusted content is collected from external platforms as described in the 'SCRAPE' section of references/reverse-engineering.md.\n
- Boundary markers: No specific delimiters or safety instructions are provided to help the agent distinguish its own core instructions from the content of the scraped posts.\n
- Capability inventory: The agent uses results to create content strategies and generate new social media posts.\n
- Sanitization: The skill lacks validation, escaping, or filtering steps for the ingested external text.\n- [NO_CODE]: The skill consists entirely of Markdown documentation and strategy guidelines without any executable scripts, binary files, or automated tool integrations.
Audit Metadata