ultralab-scanners: A Zero-Dependency TypeScript Scanning Tool Redefining SEO and AI Search Optimization

ultralab-scanners is a zero-dependency scanning library written purely in TypeScript, providing three core functions: SEO, AEO (Answer Engine Optimization), and AI prompt security auditing. Its features include: zero cost (no API calls, LLM inference, or external services required), ultra-fast execution (SEO/AEO scanning <50ms, prompt scanning <5ms), deterministic output (no AI hallucinations), privacy-first (data never leaves the local machine). It is suitable for content optimization and security needs in the AI era.

Existing SEO tools have pain points such as paid subscriptions, dependency on API keys, and the need to launch headless browsers, which increase costs and reduce efficiency. ultralab-scanners is designed based on these pain points, with core concepts: zero cost, zero dependency (pure TypeScript implementation), ultra-fast execution, deterministic output, and privacy-first.

The SEO scanner uses pure HTML parsing, no headless browser or Lighthouse required. It covers 8 categories of checks: meta tags (title length, meta description, etc.), heading structure, image optimization (alt attributes, lazy loading), link checks, social/Open Graph tags, technical specifications (HTTPS, indexability), structured data (JSON-LD), and performance tips. The output includes status (pass/warning/fail), a 0-100 score, an A-F grade, and a comprehensive summary.

AEO (Answer Engine Optimization) is an innovative feature of ultralab-scanners, which no other tools on the market have. It optimizes content to adapt to AI search engines like ChatGPT Search and Perplexity, ensuring content is cited by AI. It includes 32 checks covering 8 categories: FAQ Schema, answer-ready content, structured data richness, content clarity, citation friendliness, AI crawler access permissions, llms.txt support, and entity clarity (E-E-A-T).

Targeting LLM prompt injection threats, this scanner audits the security defense measures of system prompts, covering 12 attack vectors: role boundary escape, instruction boundary breach, lack of data protection, insufficient output control, multilingual protection, Unicode attacks, length limit bypass, indirect injection, social engineering, harmful content generation, abuse prevention, and insufficient input validation. It supports English/Traditional Chinese, implemented with regex, executes in <5ms, and outputs defense coverage rate and confidence score.

Application scenarios include: setting score thresholds in CI/CD pipelines (build fails if below threshold), batch content auditing to generate AEO health reports, LLM prompt security review (before deployment). Technical advantages: zero dependency (avoids dependency hell, size <15KB, reduces supply chain risks), ultra-fast (SEO/AEO <50ms, prompt <5ms), structured JSON output for easy integration. It has been integrated into UltraProbe (1600+ monthly active users).

Comparison with existing tools:

• Lighthouse: Requires Chrome support; this tool only needs pure HTML parsing.
• Commercial tools like Ahrefs: Require paid APIs; this tool is free.
• Screaming Frog: Needs Java runtime and high annual fee; this tool has no such requirements. The core advantage: The AEO function is unique in the market; no other open-source/commercial tools provide AI search engine optimization checks.

ultralab-scanners provides zero-cost, privacy-first optimization and security solutions with deterministic algorithms, which can match or even surpass AI solutions without using AI. The project uses the MIT license; contributions are welcome: adding new AEO checks (evolution of AI search signals), prompt defense vectors (new attack types), multilingual support, and performance benchmarking data.