The Agent Skills Directory

[SAFE]: The skill defines a structured workflow for auditing source code for injection vulnerabilities. It utilizes established security scanners and provides clear detection patterns.\n- [SAFE]: Integration with third-party tools such as Semgrep, Bandit, and Gosec is performed locally via command-line invocation for static analysis purposes.\n- [SAFE]: Documentation in references/detection-patterns.md contains examples of vulnerable code for educational identification; these are treated as data by the agent and are not executed.\n- [SAFE]: No malicious patterns such as command injection, credential harvesting, or data exfiltration were observed in the skill instructions or scripts.

injection