I Built a Tool to Stop Losing My Claude Code Conversation History

Source: Dev.to

The Problem: Lost Claude Code Conversation History

A few weeks ago I needed to revisit a debugging session. Claude had walked me through a nasty race condition in my app—it took over an hour, and the fix was subtle. I knew exactly which session it was, but when I went to find the JSONL file it was gone. No warning, no “this file will be deleted in 3 days.” Just gone.

If you’ve been using Claude Code for more than a couple of months, this has probably happened to you too. Claude Code stores conversations as JSONL files under ~/.claude/projects/, and old files are automatically deleted over time. You can disable auto‑deletion in settings, but that only solves part of the problem:

• /compact still lossy‑summarizes your context.
• Version updates can break session compatibility.
• Even with deletion disabled, JSONL files are scattered across directories with no way to search across sessions.

Existing Tools

• claude-history (Rust TUI) – great for browsing, but reads JSONL files directly, so deleted files are invisible.
• Claude Code History Viewer (desktop app) – same limitation.
• claude-mem – persists data into its own database and adds semantic search, but brings in a full Node.js/MCP server stack that’s overkill for simple archiving.

What I needed was a simple, durable archive I could set up once and forget about.

Introducing claude-vault

claude-vault is a single‑binary Rust tool that imports Claude Code conversations into an SQLite database with full‑text search. No Node.js, no Python, no MCP server—just download and run.

Importing Conversations

claude-vault import
# Imported 94562 messages (0 skipped, 12847 filtered, 0 errors) from 203 files

Once the data is in SQLite, it survives file deletion, compaction, and updates.

What Gets Imported

Claude Code JSONL files contain a lot of noise (tool results, system tags, file‑read outputs, progress messages). claude-vault strips all of that, keeping only:

• Your questions
• Claude’s responses
• Code‑modifying actions

Searching the Archive

claude-vault uses SQLite’s FTS5 with Porter stemming, so searches are flexible:

claude-vault search "race condition fix"
claude-vault search "deploy" --project my-app --since 2025-01-01

You can also get JSON output for piping back into Claude:

claude-vault search "previous auth implementation" --json

Automating the Import

Manually running import works, but I kept forgetting. The real fix is to hook the import into Claude Code’s lifecycle.

Add the following to ~/.claude/settings.json:

{
  "hooks": {
    "PreCompact": [
      {
        "hooks": [
          {
            "type": "command",
            "command": "claude-vault import >/dev/null 2>&1"
          }
        ]
      }
    ],
    "SessionEnd": [
      {
        "hooks": [
          {
            "type": "command",
            "command": "claude-vault import >/dev/null 2>&1 &"
          }
        ]
      }
    ]
  }
}

• PreCompact – captures the full conversation before /compact summarizes it.
• SessionEnd – archives the session in the background when you exit.

Once set up, every session is archived automatically without any further thought.

How It Differs from Other Solutions

It does one thing: make sure your conversations don’t disappear. Nothing more, nothing less.

Installation

cargo install claude-vault
# or download a prebuilt binary from GitHub Releases

Call to Action

Run claude-vault import now. If you’ve been using Claude Code for a while, some of your old sessions might already be gone—archive what’s left before it’s too late.

GitHub:

Have you lost Claude Code sessions you wish you could get back? What’s your approach to preserving conversation history? I’d love to hear what others are doing.