communications-specialist

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches organization details from https://wazoo.dev to synchronize content with the latest company metadata. This targets a vendor-associated domain.
  • [PROMPT_INJECTION]: The skill includes instructions to proactively monitor milestones and 'own the public narrative,' using role-play imperatives to guide agent behavior.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it transforms external technical signals, such as commit messages and JSON-LD graphs, into polished narratives without explicit sanitization.
  • Ingestion points: Technical milestones, commit messages, and the wazoo.dev JSON-LD graph.
  • Boundary markers: Absent; there are no specific delimiters defined to separate instructions from technical data.
  • Capability inventory: Documentation modification (Mintlify MDX) and social media account management.
  • Sanitization: No explicit validation or escaping of ingested technical signals is provided in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 11:52 PM