Answer one question well: “What changed in my logs?”

Built after too many incidents where logging costs exploded and no one knew why.

Logging failures don’t look like failures. They look like:

• debug logs left on
• retry loops
• health checks spamming
• one service going insane

This tool shows what actually changed between two log snapshots.

Compare two log files to find spikes and new patterns.

python log-xray.py --base baseline.log --new incident.log

Output: An HTML report named log_report.html showing:

• Total volume % change
• Top "spikes" (existing logs that exploded)
• New signatures (patterns that didn't exist before)

Scan a single file to find repetitive noise.

python log-xray.py app.log

Output: An HTML report showing entropy scores and noise reduction candidates.

• Runs 100% locally
• No uploads
• No SaaS
• No tracking

Your logs never leave your machine.

See examples/report.html for a sample report.

MIT