ai-data-extractor

Why we included this project

AI coding assistants each keep your chat history in their own private storage, and this tool pulls that data out into a single normalized JSONL file. It auto-discovers the local data folders for ten different assistants, reading SQLite, JSON, and JSONL sources read-only, and preserves code context, diffs, tool calls, and timestamps wherever the source tool actually records them. That makes it practical for building fine-tuning datasets from your own real sessions, running personal analytics on how you actually work, or backing up years of conversations before an app clears its local database. The search covers macOS, Linux, and Windows conventions automatically, so you don't have to tell it where anything lives, and each per-source extractor can run standalone when you only need one tool. Teams writing their own extractors will also find the Cline and Aider implementations useful as worked examples of parsing very different storage shapes.