detecting-sql-injection-vulnerabilities

The skill is internally consistent and not overtly malicious: it reads local code, analyzes SQL injection risks, and writes reports. However, it is a high-risk security-audit capability for an AI agent because it combines untrusted code analysis with Bash execution and write access, so it should be treated as a powerful but legitimate offensive-security-adjacent tool rather than benign general-purpose documentation.