Logo
Initializing AI Engine

Technical Architecture

A deep dive into the modular, secure, and high-performance core of SemVer AI Tool.

Layered System Design

1. CLI Router (src/cli.sh)

The entry point that orchestrates command delegation. It dynamically calculates paths and routes execution to specialized controllers while maintaining environment isolation.

2. Orchestration Layer (src/commands/)

Controllers like release.sh and init.sh that manage the sequential flow of operations, from config loading to AI synthesis and git persistence.

3. Core Services (src/core/)

The engine room: ai_client for LLM communication, git_manager for history parsing, and version_manager for semantic logic.

Intelligence Module

The ai_client.sh module bridges the gap between your code and the LLM. It constructs complex prompts that combine your technical context with specialized documentation templates.

  • Powered by Groq Cloud (LLaMA 3.3)
  • Zero-latency (typically <1s responses)
  • Native Markdown synthesis

Data Integrity & Persistence

Secure Config

Uses node -e for safe JSON parsing, bypassing typical shell injection vulnerabilities when reading local keys.

Git Atomic Ops

Ensures package.json updates, documentation generation, and tagging happen in a single, rollback-safe sequence.