Can I run CQ-AI scans in parallel?

Yes, CQ-AI features a parallel scanning architecture that distributes work across multiple threads while maintaining a proof of order-independence for the results.

What makes CQ-AI scanning deterministic?

CQ-AI uses SplitMix64 seeding to ensure that providing the same seed and codebase always produces identical findings, eliminating the randomness often found in complex security tools.

How does ternary polarity classification work?

It maps security findings to a balanced ternary system (GF(3)), where findings are classified as +1 (Critical), 0 (Medium), or -1 (Info) to streamline remediation efforts.

Does this skill work with Claude's MCP?

Yes, CQ-AI integrates as an MCP tool, allowing Claude to intelligently configure scans, interpret findings, and prioritize the most critical security risks.

CQ-AI Deterministic Security Scanner

Name: CQ-AI Deterministic Security Scanner
Author: plurigrid

byplurigrid

•

セキュリティとテスト

Performs reproducible, AI-enhanced security audits using deterministic analysis and ternary polarity classification.

CQ-AI is a specialized security skill for Claude Code that extends the Code Query (CQ) framework with a deterministic analysis engine. By utilizing SplitMix64 seeding and balanced ternary classification (GF(3)), it guarantees consistent vulnerability findings across different runs, environments, and team members. This skill is essential for security-focused developers who require mathematical proofs of scan consistency and AI-guided prioritization of critical risks like SQL injection, RCE, and authentication bypasses.