Security-First Code Review

AI + Security Tools.
Zero Blind Spots.

diffray combines industry-leading security scanners (TruffleHog, Semgrep) with AI interpretation. Catch leaked secrets and vulnerabilities before they reach production.

Why AI Alone Isn't Enough

AI models are great at reasoning, but specialized tools catch what AI misses

AI-Only
Context awareness
Reasoning about patterns
May miss edge cases
False negatives possible
Tools-Only
Precise detection
Low false negatives
High false positives
No context understanding
AI + Tools (diffray)
Higher detection rates
Lower false positives
Better context
Actionable remediation

Tools detect. AI validates and explains. You get actionable results.

TruffleHog
Secrets Detection

TruffleHog Integration

Industry-leading secrets scanner that detects leaked credentials, API keys, and sensitive data before they reach your repository.

What It Detects:

AWS access keys
GitHub tokens
Private keys (RSA, SSH)
Database connections
700+ API key formats
JWT tokens
Cloud credentials
Session secrets

How TruffleHog Works:

1

Regex Patterns

Matches known credential formats

2

Entropy Analysis

Detects high-entropy strings that look like secrets

3

Verification

Attempts to verify if detected secrets are active

AI Enhancement

  • Validatesconfirms it's a real credential, not a false positive
  • Assesses severitytest key vs production credential
  • Provides contextwhere it's used and potential impact
  • Suggests remediationhow to rotate and secure

What Semgrep Detects:

Injection Vulnerabilities

SQL injection, command injection, XSS

Auth Issues

Hardcoded passwords, weak crypto

Insecure Configs

Debug mode, CORS misconfigurations

Code Quality

Null pointer, resource leaks

30+ Languages Supported:

TypeScriptJavaScriptPythonGoJavaKotlinRubyPHPC/C++RustSwift
Semgrep
Static Analysis

Semgrep Integration

Fast, open-source static analysis that finds bugs and security vulnerabilities using semantic pattern matching. Unlike regex tools, Semgrep understands code structure.

AI Enhancement

  • Validates relevancechecks if finding applies to your code
  • Reduces noisefilters false positives based on context
  • Prioritizes severityranks by actual risk
  • Explains vulnerabilityhow it could be exploited
  • Provides fixesspecific code changes

The Security Pipeline

How security tools integrate into every code review

1

PR Changes Detected

New code is pushed to a pull request

2

TruffleHog Scans

Secrets scanner checks only new changes

3

Semgrep Analyzes

Static analysis runs on modified code

4

AI Receives Context

Tool output + code context sent to AI

5

AI Validates & Filters

False positives removed, findings enriched

6

Actionable Report

Clear, prioritized issues with fixes

Differential Scanning

Only analyzes changes in the PR — fast scans, no noise from existing issues

AI Validation

Every finding reviewed by AI — false positives filtered, context added

Unified Reporting

All findings in one organized comment with consistent formatting

Catch Security Issues
Before Production

Leaked secrets and vulnerabilities don't stand a chance. Start securing your code today.

TruffleHog + Semgrep included
No configuration needed
Runs on every PR