basercms5-development

Pass

Audited by Gen Agent Trust Hub on Jul 5, 2026

Risk Level: SAFEDATA_EXFILTRATION
Full Analysis
  • [DATA_EXFILTRATION]: The skill instructs the agent to read sensitive environment and installation configuration files for discovery and investigation purposes.
  • Evidence: The skill specifies reading htdocs/config/install.php, htdocs/config/.env.php, and htdocs/config/.env to confirm database connection settings and routing/debug configurations.
  • Context: While common for development tasks, reading these files exposes sensitive information (credentials, secret keys) to the agent context.
  • [DATA_EXFILTRATION]: Ingestion of untrusted or sensitive data for environment discovery (Category 8).
  • Ingestion points: htdocs/config/install.php, htdocs/config/.env.php, and htdocs/config/.env.
  • Boundary markers: Absent in this guideline file.
  • Capability inventory: The skill mentions logging through CakeLog and procedural PHP implementations, though no direct network exfiltration commands are present in the skill body.
  • Sanitization: Not defined for the ingested configuration data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 5, 2026, 05:43 AM
Security Audit — agent-trust-hub — basercms5-development