Is secret_leak turned on by default?

No, not out of the box. secret_leak ships off. duration_drift, cost_drift, and tool_error are the three built-in classifiers that ship on; the rest, including secret_leak, you turn on per project once you want them checking your traces.

That’s a deliberate default rather than an oversight. A new project’s traffic patterns are unknown, and a classifier running before anyone has looked at what a project’s traces normally contain would just be alerts nobody’s ready to act on. Turning it on is a choice you make, not something that starts happening to your data the moment you connect a trace.

Once it’s on, it runs the same way every other built-in does: every trace, one narrow check, a finding when it fires, and triage before anything reaches a person.

keep reading

More on this.

Can I see the actual secret Tessary flagged?No. Detections carry redacted evidence, a pattern name and a short snippet, never the credential itself.answer →Does mentioning a password in conversation trigger secret_leak?No. It's anchored on the shape of a live credential, like a key prefix or private-key header, not the word 'password' in ordinary prose.answer →Does secret_leak see credentials in inner tool calls, or only the final response?Everything. It checks every span in a trace, including inner tool calls, not just the message the user ends up reading.answer →Everything on tessary secret leakThe definition, and every question answered under tessary secret leak.concept →

Send us the traces you already emit.

book a demo