memory-publication-auditor

Pass

Audited by Gen Agent Trust Hub on May 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates locally to enhance privacy by identifying sensitive data in files and logs. It uses standard agent tools for its intended purpose and does not attempt network communication or unauthorized access to system resources.\n- [PROMPT_INJECTION]: The skill processes untrusted local data, creating a surface for indirect prompt injection. This risk is managed through deterministic scanning and the use of structured audit templates to separate data from instructions.\n
  • Ingestion points: Reads local files and logs from user-specified paths via the --input argument.\n
  • Boundary markers: Scanned content is isolated within Markdown code blocks and tables in the audit report template.\n
  • Capability inventory: Includes Bash to execute the local scanner script, alongside Read, Write, Edit, and Glob for file operations.\n
  • Sanitization: The scanner script automatically redacts detected sensitive strings (emails, IPs, keys) and escapes HTML characters in evidence snippets before the agent reviews them.
Audit Metadata
Risk Level
SAFE
Analyzed
May 10, 2026, 08:54 PM