lextrackai

AI Compliance as Code Leger is a next-generation, automated CLI tool designed specifically for AI systems. It scans a project's codebase for AI-related files—including model files, configuration files, external API calls, and prompt engineering data—generates a detailed, human-readable compliance report, and computes a SHA-256 hash of the report for secure, tamper-proof record keeping.

The vision behind lextrackai is to simplify and automate compliance for AI projects by capturing essential AI artifacts quickly and reliably, setting the foundation for future blockchain integration and advanced content analysis.

Features

• Recursive Directory Scanning:
  Traverses directories to identify files based on extensions like .h5, .pt, .pkl, .json, .yaml, .yml, .env, .js, .py, etc.

• Dynamic Content Analysis:
  Analyzes text-based files to detect:

  • External API Calls: Uses a dynamic list of regex patterns to detect API calls from vendors such as OpenAI, Google Gemini, Anthropic, etc.
  • Prompt Declarations: Detects variables named prompt or promptTemplate which indicate prompt engineering.

• Human-Readable Compliance Report:
  Generates a detailed report including:

  • Total number of scanned files.
  • A list of detected files with their paths, extensions, and any specific detections.

• SHA-256 Hash Computation:
  Computes a secure SHA-256 hash of the generated report to ensure its integrity.

• Simple CLI Interface:
  Easily run the tool using:

  npx lextrackai scan