Read-only by default
TraceX connects to your databases with read-only access and runs only read-only queries. Nothing in production is changed.
Powerful by design. Controlled by you.
TraceX was built to investigate production systems without changing them, and without keeping your data.
Production writes
0
read-only credentials and read-only queries only
Indexes kept
0
TraceX reads live; it keeps no copy of your code or data
Export lifetime
24h
until spreadsheets, CSV and PDF exports are deleted
Models per answer
2
one writes the answer, a second checks it against the evidence
Read-only
Nothing in production is changed.
Database connections use read-only credentials, and every query is checked before it runs. Anything that would write is refused.
- Read-only credentials for every database
- Write and schema queries refused before they run
- Code changes only as pull requests you review
SELECT id, state FROM refunds WHERE order_id = 48213UPDATE refunds SET state = 'paid'DROP TABLE refundsOnly read-only queries run. Everything else is refused before it reaches your database.
Evidence, not guesses
Every root cause and answer cites the records, code and log lines it came from, and a second model checks the answer against them.
Your records aren't kept
No index of your code or data. Answers keep what was looked up and where, not the rows; exported files are deleted after 24 hours.
You approve every change
Fixes arrive as pull requests for your review. In your terminal, TraceX asks before every edit and command.
Enterprise sign-in
SSO, two-factor sign-in, roles for admins, investigators and viewers, revocable sessions and a full audit log.
Your own model keys
Run TraceX on your own Anthropic or Gemini account, so AI usage is billed to you, never through TraceX.
Details
How your data is handled.
Does TraceX store my customers' data?
No. TraceX reads your systems live when it investigates or answers. It keeps what it looked up and where (for example "34 rows in refunds"), not the rows themselves; evidence excerpts are removed when an investigation finishes. There is no index of your code or database.
Can TraceX change my production data?
No. Database connections use read-only credentials and TraceX only runs read-only queries. Code changes are pull requests that your team reviews and merges.
What happens to files I export?
Spreadsheets, CSV and PDF files you ask for are deleted after 24 hours. After that their download links stop working.
Who can see what?
Admins, investigators and viewers have different permissions. Sign-in supports SSO and two-factor authentication, every session can be revoked, and actions are recorded in an audit log.
Is my data used to train models?
TraceX never trains models on your data. You can also run TraceX on your own Anthropic or Gemini account, so your AI traffic goes through your own agreement with the provider.
What can the CLI do on my machine?
In coding mode the CLI can read, edit and run commands only inside the repository it's started in, and it asks for your approval before every edit and command. Your sign-in is stored in a file only you can read, and can be revoked from the dashboard.
Find the root cause.
Ship the fix.
TraceX is free during the private beta. Connect a database and a repository and ask your first question in minutes.