behavioural-analytics

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of purely instructional Markdown content. No security vulnerabilities or malicious patterns were identified in either the skill definition or the evaluation cases.
  • [PROMPT_INJECTION]: The instructions are consistently focused on data analysis methodologies and do not attempt to override agent safety filters or system instructions.
  • [DATA_EXFILTRATION]: There is no evidence of unauthorized data access or transmission. No sensitive file paths, hardcoded secrets, or network calls are present.
  • [REMOTE_CODE_EXECUTION]: The skill does not include or reference any external scripts, package managers, or dynamic execution capabilities.
  • [OBFUSCATION]: A thorough check for Base64, zero-width characters, and homoglyphs returned no results. All content is transparent and readable.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 03:38 PM
Security Audit — agent-trust-hub — behavioural-analytics