social-trend-monitor
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to monitor and process content from external platforms including Reddit, Instagram, TikTok, and Twitter. This ingestion of untrusted external content exposes the agent to indirect prompt injection vulnerabilities.
- Ingestion points: External social media content from Reddit, Instagram, TikTok, and Twitter as defined in the skill triggers and description.
- Boundary markers: Absent; no instructions are provided to delimit external data or warn the agent to ignore instructions embedded within the ingested content.
- Capability inventory: No tools, scripts, or subprocesses are defined in the manifest, but the skill is intended to generate trend analysis reports.
- Sanitization: No sanitization, validation, or filtering logic is specified for the external data.
- [NO_CODE]: The skill package contains only documentation and manifest metadata. No executable scripts, binaries, or configuration files were found in the provided files.
Audit Metadata