community-engagement
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface.
- Ingestion points: Uses WebSearch and WebFetch to read content from Reddit, Facebook, and Discord threads (untrusted external data).
- Boundary markers: None explicitly defined to separate external data from instructions in SKILL.md.
- Capability inventory: Generates community monitoring reports and integrates with lead-qualification and linkedin-outreach skills.
- Sanitization: No explicit sanitization or filtering of fetched content is mentioned.
- [COMMAND_EXECUTION]: No shell commands or script executions were found. The skill relies on natural language instructions for the agent to follow.
- [SAFE]: No hardcoded credentials, obfuscation, or persistence mechanisms were detected in the skill definition.
Audit Metadata