ObfusTrace

Debug-equivalent synthetic traces for safe AI and support sharing.

ObfusTrace turns a production error, stack trace, log excerpt, or configuration fragment into a safe synthetic twin before it is pasted into AI, Jira, Slack, GitHub, or vendor support.

It does more than replace values with [REDACTED]:

• repeated originals receive stable aliases;
• IPs, ARNs, UUIDs, emails, URLs, JWTs, and credentials retain useful structure;
• line count and stack-frame count remain unchanged;
• a structural fingerprint proves diagnostic equivalence;
• a second scan must find zero residual sensitive values before Copy is enabled;
• AI responses containing aliases can be rehydrated locally.

Raw traces and the mapping vault never leave the browser.

Run locally

npm install
npm run dev

Open http://localhost:3000.

Demo

1. Inspect the included synthetic production failure.
2. Click Create safe trace twin.
3. Review detected categories and stable synthetic aliases.
4. Confirm the sanitization certificate is CLEAN.
5. Copy the safe trace for external debugging.
6. Paste an AI response containing aliases and click Rehydrate locally.

Sample protection

The bundled trace includes a synthetic JWT, Bearer header, AWS access key, AWS ARN/account, database username/password, customer email, UUID, internal hostname, partner hostname, and repeated private IP.

Expected output:

• 10 unique sensitive values protected;
• repeated IP maps consistently to 192.0.2.1;
• email maps to user1@example.invalid;
• stack frames and line count remain unchanged;
• residual findings equal zero;
• copy gate becomes enabled;
• local rehydration restores original values.

Quality gates

npm run lint
npm run build
npm run test:e2e

Playwright validates detection, stable aliasing, residual scanning, diagnostic fingerprint preservation, copy gating, reversible local mapping, and mobile behavior.

Boundaries

ObfusTrace reduces accidental disclosure risk; it is not a guarantee that arbitrary text contains no sensitive business information. High-risk organizations should combine it with existing secret scanning, DLP policy, and human review. The sample data is synthetic.