anibon-donation-classifier

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill is a purely instructional markdown document intended to guide text classification and tone selection for donation timestamps. It does not define any tools, perform network operations, or access sensitive system files.
  • [NO_CODE]: No executable scripts, binaries, or package manifests (e.g., requirements.txt, package.json) are included in the skill.
  • [PROMPT_INJECTION]: Evaluation of the 'Iron Rules' and 'Rationalization Red Flags' confirms they are benign instructional constraints aimed at ensuring output consistency. No malicious patterns, jailbreak attempts, or safety bypass instructions were found.
  • [SAFE]: The skill explicitly instructs the agent to avoid quoting offensive content verbatim, reflecting a safety-conscious design for processing user-generated input.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 01:30 PM
Security Audit — agent-trust-hub — anibon-donation-classifier