skills/mehdic/bazinga/security-scan/Gen Agent Trust Hub

security-scan

Pass

Audited by Gen Agent Trust Hub on Apr 26, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches and installs various security scanning tools (including bandit, semgrep, gosec, and brakeman) from well-known package registries such as PyPI, NPM, RubyGems, and GitHub if they are not already installed in the environment.- [COMMAND_EXECUTION]: Executes local shell and PowerShell scripts to perform language detection, tool coordination, and database management. It also executes internal Python scripts to track session data and save results to a local SQLite database.- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it processes and summarizes output generated by external security scanners which could contain malicious content designed to influence the agent.
  • Ingestion points: Reads from tool-generated reports at bazinga/artifacts/{SESSION_ID}/skills/security_scan.json.
  • Boundary markers: Absent; the instructions do not specify delimiters or instructions to ignore embedded commands within scanner findings.
  • Capability inventory: Extensive shell command execution via the Bash tool and file system access via the Read tool.
  • Sanitization: The skill parses structured JSON data but interpolates raw findings (titles/descriptions) directly into the final summary report provided to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 26, 2026, 08:23 PM
Security Audit — agent-trust-hub — security-scan