Audit code contexts to identify and report exploitability findings.
npx skills add https://github.com/yunmengya/PHP_AUDIT_SKILLS --skill ai-audit-mcp
Please help me install this Agent Skill. Skill: ai-audit-mcp Source: https://github.com/yunmengya/PHP_AUDIT_SKILLS/tree/main/ai-audit-mcp Command: npx skills add https://github.com/yunmengya/PHP_AUDIT_SKILLS --skill ai-audit-mcp
This Skill automates the process of conducting in-depth security audits using AI, transforming raw code context into actionable exploitability findings.
Run the ai-audit-mcp skill to audit the provided AI audit context for project 'my-php-app'.
High-intent search queries and answers about installing and using this skill.
AI-driven vulnerability analysis identifies exploitability by processing detailed code context, including routes, call graphs, and function bodies, to generate normalized reports on sources, sinks, and taint for PHP vulnerabilities like SQL injection or RCE.
To perform an AI security audit, you need specific context JSON files that contain detailed code context, including routes, call graphs, function bodies, and file contents, which the AI analyzes to identify potential security vulnerabilities.
Yes, you need a local Claude CLI to run AI security audits, as the Skill leverages Claude models to deeply analyze the provided code context and generate normalized vulnerability reports with proof-of-concept details.
You audit code context by running the Skill against your prepared context JSON files; the AI analyzes the provided routes, call graphs, and file contents to output normalized findings detailing source, sink, taint, and proof-of-concept exploitability.
The best way to report SQL injection or RCE vulnerabilities is by generating normalized findings that detail the source, sink, taint, and proof-of-concept, which AI auditing automatically produces from the analyzed code context.
No, this AI audit approach requires a pre-existing static analysis context; it functions by deeply auditing specific context JSON files generated from a prior analysis to identify and detail potential code vulnerabilities.
rainoftimeBuild call graphs and analyze programs across function boundaries.
nodnarbnitramAutomate SOC II triage from Linear tickets to commits and PRs.
freire19Develop proof-of-concept exploit plans from vulnerability findings and CVE research.
allanninalGenerate custom Semgrep rules for security vulnerability detection and code pattern enforcement.
