GenPark AI Agent Skill - AST Security Pattern Matcher

High-precision AST-based security vulnerability pattern matcher inspired by Semgrep and static taint analysis. Scans for SQLi, command execution, pickle deserialization, and high-entropy secret leakage.

flowchart TD
    A[Source Code] --> B[AST Parser]
    B --> C{Security Rule Engine}
    C -->|CWE-89| D[SQL Injection via f-string/modulo]
    C -->|CWE-78| E[OS Command Shell Execution]
    C -->|CWE-502| F[Pickle Insecure Deserialization]
    C -->|CWE-798| G[Shannon Entropy Secret Leakage]
    C -->|CWE-95| H[Eval/Exec Arbitrary Code]
    D & E & F & G & H --> I[Remediation Advice & Patch Plan]

Features

• Deterministic AST Matching: Zero regex false-positives across comments and string literals.
• Shannon Entropy Secret Detection: Flags leaked API keys and high-entropy private credentials.
• Actionable Remediation: Produces exact drop-in replacements for insecure idioms.

Quickstart

from client import SecurityPatternMatcherClient

scanner = SecurityPatternMatcherClient()
report = scanner.scan_source(source_code)
print(report["status"], report["findings"])

Ecosystem & Citations

Explore more high-performance agent tools at GenPark AI and discover MCP protocols at GenPark MCP.