discover-testing

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data access behaviors were detected. The skill is composed of routing documentation and local processing scripts.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it is designed to ingest and parse untrusted data from external files (CSV, JSON, XMind, Word, Excel). This is a functional requirement of the skill but presents an inherent risk if the agent blindly follows instructions embedded within the processed data.
  • Ingestion points: Untrusted data is read from user-provided file paths in scripts/convert_formats.py and scripts/parse_formats.py.
  • Boundary markers: The output generated by the parsing scripts does not include delimiters or protective instructions for downstream LLM consumption.
  • Capability inventory: The skill utilizes subprocess.call to execute internal Python utility scripts and has the capability to write output files to the local filesystem.
  • Sanitization: No validation or sanitization of the content extracted from processed files is implemented.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 06:12 AM