security-expert

To audit code for prompt injection and OWASP Top 10 vulnerabilities, use this skill to review source files and parameters. It analyzes code against security protocols, enforcing path validation and parameterized queries to identify and mitigate malicious execution risks.

MCP-specific security threats in AI coding environments include prompt injection, data exfiltration, and tool poisoning. These vulnerabilities exploit AI-assisted development tools to execute malicious code or leak sensitive data without strict output structuring and validation.

To secure MCP tools before deployment, run a security audit in review mode to inspect parameters and structured outputs. This ensures the tool adheres to strict security protocols by enforcing path validation and parameterized queries against potential data leakage.

Yes, you can use automated security analysis for AI-assisted development tools. This skill provides expert security analysis focusing on path validation and structured output to prevent data exfiltration and guide secure implementation practices by default.

Yes, the security review strictly enforces parameterized queries and path validation. It requires adherence to these secure coding practices to prevent malicious code execution and data leakage within the analyzed AI coding environment and MCP ecosystem.