Secrets used by reference

The idea

Passwords, codes and captured tokens are stored as references; the platform fills or stages them into the guest and masks output, so no value reaches events, logs or model context.

The problem
A worker using accounts would otherwise see and leak its credentials.
Why it matters
The worker uses credentials without holding their values in context.
Where it stands
Built, off in production as of

Recorded evidence

  • Local run on the real backend · Every secret by reference

A step-by-step replay of this idea is not published yet. The library adds one when a recorded run shows it.

  • Worker's own mailbox

    A worker gets its own email account to sign up, confirm and sign in to sites by itself, with mail treated as outside content.

  • Device-code sign-in for the worker's tools

    An admin signs the worker's gh or Codex in on its computer with the vendor's device code, which reaches only the admin and is never stored; the doctor signs a wrong account out.

  • Redaction on write and the credential vault

    Secret-shaped values are removed before events, steps, tool results and model context are stored, and customer credentials are sealed with AES-256-GCM bound to org and credential.

Turn the numbers into shipped work.

Crewshal runs these choices for you: a persistent AI worker with memory and rules, on your Claude and Codex subscriptions.