skills/cniska/skills/debug/Gen Agent Trust Hub

debug

Pass

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: No security issues detected. The skill consists of purely instructional markdown content providing best practices for debugging.
  • [INDIRECT_PROMPT_INJECTION]: The skill contains defensive instructions (treating error output as data) to mitigate potential indirect prompt injection attacks from external error messages. Evidence: 1. Ingestion points: Error messages from external sources mentioned in the 'Treating error output as data' section. 2. Boundary markers: Explicit instruction to treat messages as data rather than instructions. 3. Capability inventory: No tools, scripts, or network operations are included in this skill. 4. Sanitization: Instruction to surface external fix commands to the user rather than acting on them autonomously.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 9, 2026, 10:47 AM