Show HN: Built Open Source AI Coding Assistant Which Rivals to Cursor (www.lynelabs.com)

Lynecode is an open-source, terminal-first AI coding assistant—positioned as a rival to Cursor—that lets developers chat with and edit any project folder from the command line. It’s free to use (you bring your own API keys), installs via pip (Python 3.9+), and runs on Windows, macOS, and Linux. Key user-focused features include project-wide indexing for context-aware suggestions, inline file/folder attachments, automatic local safety snapshots before AI edits (recoverable diffs even without git), and strict local-only data handling. The UI emphasizes a fast chat workflow with commands for creating files/folders, inserting or replacing code blocks, AST-aware search, Semgrep scans, lint checks, git diffs, and an embedded terminal. For the AI/ML and dev communities, Lynecode matters because it combines model flexibility, privacy, and workflow safety in a compact, open toolchain. It supports OpenAI (gpt-4.1/gpt-5), Google Gemini (gemini-2.5/pro), and Azure OpenAI—models can be switched at runtime—so users can optimize latency, cost, or capabilities and even leverage free provider credits. Being fully open-source and locally focused enables auditability, customization, and integration into CI/dev workflows without IDE lock‑in. Technical implications include safer experimentation with generative edits, easier reproducibility through indexed context, and a lightweight path for teams or researchers to evaluate multiple LLM providers in real projects.