skills/simota/agent-skills/bond/Gen Agent Trust Hub

bond

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely composed of instructional markdown files and reference templates. It does not include executable scripts, shell commands, or external code dependencies.
  • [DATA_EXPOSURE]: No evidence of hardcoded credentials, sensitive file access, or unauthorized network operations was found. The skill defines data flows for metrics and feedback but does not implement exfiltration mechanisms.
  • [REMOTE_CODE_EXECUTION]: The skill does not perform any remote package installations or script executions. It relies on internal logic and references to guide the agent's behavior.
  • [PROMPT_INJECTION]: There are no patterns suggesting attempts to bypass safety filters or override system-level instructions. The instructions are focused on business logic and user retention strategies.
  • [OBFUSCATION]: All content is provided in clear text with standard markdown formatting. No encoded strings, hidden characters, or homoglyph substitutions were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 09:28 AM
Security Audit — agent-trust-hub — bond