content-repurposer

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from user-supplied sources, which could contain malicious instructions designed to influence the agent's behavior.
  • Ingestion points: The skill reads input from file paths, external URLs (via WebFetch), or pasted text during the source loading step.
  • Boundary markers: The instructions do not define clear delimiters for source content or provide a system-level instruction to ignore potential commands embedded within the ingested data.
  • Capability inventory: The skill possesses the ability to read internal knowledge files (e.g., brand voice and personas), write derivative content to the local file system (output/repurposed/), and call other functional skills like linkedin-post and content-writer.
  • Sanitization: While the skill emphasizes factual accuracy and prevents fabrication, it lacks a dedicated mechanism to sanitize or filter out non-data instructions from the source material.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 06:28 AM
Security Audit — agent-trust-hub — content-repurposer