Engineering Shiply: Parsing AST Diffs for Semantic Git Commits & Developer Telemetry

## The Problem: The Git Commit Burden Every developer knows the sensation: you have just spent three hours solving a nasty race condition across four microservices. Your tests pass. You are ready to push. And then you face the empty commit prompt: `git commit -m "..."` What follows is either a meaningless `git commit -m "fixed stuff"` or a painful pause to re-read your diff and write a conventional commit manually. ### Architecture of Shiply Shiply was built to solve this exact bottleneck. When you run `shiply commit`, the CLI executes: 1. **Local Diff Extraction**: It scans the git index for staged changes and summarizes the AST (Abstract Syntax Tree) modifications. 2. **Context Window Optimization**: Instead of piping thousands of lines of code into an LLM, Shiply isolates exported symbol changes, function signature diffs, and test assertions. 3. **Semantic Intent Deduction**: The model classifies the changeset into conventional standards (`feat`, `fix`, `refactor`, `perf`, `chore`) and outputs a concise explanation of *why* the change was made. 4. **Streak & Telemetry**: Updates your local activity database and increments your continuous shipping streak. Try it out on [Shiply.space](https://shiply.space)!