security-agent

Security Agent

Quick Reference — see parent for full agent ecosystem.

The Security Agent scans code changes, endpoints, and configurations for vulnerabilities before they reach production. It combines static analysis (Semgrep, CodeQL), secret scanning (Gitleaks, TruffleHog), dependency auditing, and dynamic probing to surface findings ranked by severity with evidence and fix recommendations. Its adversarial mindset assumes every input is malicious and every exposed endpoint is an attack surface.

When Not to Use

• Simple or one-off tasks — if the task is straightforward, direct execution is faster than structured methodology.
• Already established workflows — follow existing team conventions rather than introducing new frameworks.
• When automation overhead exceeds benefit — for very small scopes, the setup cost may not be justified.

Dependencies

• Python 3.8+ or Node.js 18+
• Access to relevant APIs/services for your specific use case
• Basic understanding of the domain concepts

Commands

# Refer to the skill's usage section for specific commands
# Adapt these to your workflow

Key Responsibilities

• Static vulnerability scanning: Run SAST rules for injection (SQL, command, template), XSS, SSRF, insecure deserialization, auth bypass, and cryptography misuse across every changed file
• Secret detection: Scan diffs, commit history, and config files for hardcoded credentials, API keys, tokens, and private keys — including encoded/obfuscated secrets
• Dependency audit: Check for known CVEs in direct and transitive dependencies; flag supply chain risks from typosquatting, abandoned packages, and suspicious maintenance patterns

Code Example

"""Minimal security agent pattern — scan a diff for vulnerabilities."""

import json, sys, re

def scan_diff(diff_text: str) -> dict:
    findings = []
    lines = diff_text.split("\n")

    patterns = {
        "P1": {
            "eval": r"\beval\s*\(",
            "exec": r"\bexec\s*\(",
            "raw_sql": r"\.execute\(.*['\"].*SELECT|INSERT|UPDATE|DELETE",
            "hardcoded_key": r"(?:sk-|pk-|AKIA|-----BEGIN (?:RSA |EC )?PRIVATE KEY-----)",
        },
        "P2": {
            "pickle_load": r"pickle\.loads?\(",
            "assert_true": r"assert True",
            "debug_endpoint": r"@app\.route\(.*['\"]/debug",
            "insecure_hash": r"hashlib\.md5|hashlib\.sha1",
        }
    }

    for severity, checks in patterns.items():
        for name, pattern in checks.items():
            for i, line in enumerate(lines):
                if line.startswith("+") and re.search(pattern, line):
                    findings.append({
                        "file": "changed_file", "line": i,
                        "severity": severity, "type": name,
                        "finding": f"Potential {name} detected",
                        "recommendation": "See OWASP cheat sheet for safe alternatives"
                    })

    return {
        "findings": findings,
        "summary": {
            "P1": len([f for f in findings if f["severity"] == "P1"]),
            "P2": len([f for f in findings if f["severity"] == "P2"]),
            "P3": len([f for f in findings if f["severity"] == "P3"])
        },
        "verdict": "blocked" if any(f["severity"] == "P1" for f in findings) else "needs_review" if findings else "clean"
    }

if __name__ == "__main__":
    diff = sys.stdin.read()
    result = scan_diff(diff)
    print(json.dumps(result, indent=2))