bkit-rules

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a set of methodology rules (PDCA) and quality standards for AI agent behavior within the bkit ecosystem.
  • [COMMAND_EXECUTION]: The skill provides documentation and configuration examples for defining tool permissions in the environment (e.g., wildcard permissions for npm, node, and git). These are instructions for user-level configuration and do not involve autonomous execution of dangerous commands.
  • [DATA_EXFILTRATION]: Analysis of the level detection logic shows it relies on identifying common project files (e.g., terraform, turbo.json, supabase/) to categorize project complexity. No patterns for accessing sensitive files or exfiltrating data to external domains were found.
  • [REMOTE_CODE_EXECUTION]: The skill mentions hooks (e.g., pre-write.js) and imports local templates, but does not contain instructions to download or execute code from untrusted remote sources.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 11:53 AM
Security Audit — agent-trust-hub — bkit-rules