Claude Code History Viewer

Generates an HTML page + JSON data file to browse past Claude Code conversations in a browser.

How it works

1. Run the Python extraction script to parse all JSONL session files
2. Output: claude-history-data.json (conversation data) + claude-history-viewer.html (UI)
3. Serve via python3 -m http.server 8765 and open in browser

Step 1: Generate data

Run the extraction script:

python3 ~/.claude/skills/history-viewer/extract.py

This reads all ~/.claude/projects/-Users-chethanbhatbs/*.jsonl files and produces ~/claude-history-data.json.

Step 2: Ensure viewer HTML exists

The viewer HTML is at ~/claude-history-viewer.html. If it doesn't exist, copy from the skill:

cp ~/.claude/skills/history-viewer/viewer.html ~/claude-history-viewer.html

Step 3: Serve and open

cd ~ && python3 -m http.server 8765 &
open http://localhost:8765/claude-history-viewer.html

Arguments

• --refresh — Force regenerate data even if claude-history-data.json already exists
• No args — Only regenerate if data file is missing or older than 1 hour

Notes

• Data file can be large (4-5MB for ~40 sessions). Extraction takes 5-10 seconds.
• The viewer fetches claude-history-data.json via HTTP, so must be served (not opened as file://).
• Tool calls and tool results are excluded from the viewer to keep it readable. Only human text and Claude text responses are shown.