Anyone who has spent months working with AI coding assistants knows the problem: thousands of past conversations buried in JSONL files with no way to find that one prompt where you solved a tricky database migration.
AI-Hist is a new open-source tool that pulls conversation history from Claude Code, Codex CLI, and Agent Relay into a single SQLite database, then layers FTS5 full-text search (SQLite's built-in search engine that indexes every word for near-instant lookups) on top.
The workflow is straightforward: search across all your prompts, drill into the full conversation, see the session context, and pick up where you left off. It supports incremental sync, meaning re-runs only read new bytes rather than re-importing everything from scratch.
For heavy AI coding tool users who have accumulated tens of thousands of conversations, this solves a real pain point. Those history files contain working solutions, debugging approaches, and architectural decisions that are genuinely useful months later - but only if you can actually find them. Right now, most people either grep through raw JSONL or just re-ask the same questions.
The SQLite approach keeps things simple and portable. No server to run, no external dependencies beyond Python. Your conversation history stays local, which matters when those logs contain proprietary code.