Skip to main content

​
Auth & Login

Webrain handles authenticated scraping with an encrypted local vault and a real-Chrome login flow. Secrets are decrypted in-process and injected via CDP — the value never passes through the model.

​
The credential vault

Store credentials with hidden prompts (never typed into logs):
The vault is encrypted with AES-256-GCM and supports optional TOTP seeds for 2FA.

​
Automatic login

webrain_login decrypts the secret in-process and injects it via CDP:
  • Auto-discovers the login fields and submits.
  • On a 2FA/approval gate it TOTP-injects if a seed is stored and returns waiting_for_human: true — the human acts in the headed browser, then call login again.
  • Reply is status-only. Optional port targets a specific Chrome.
  • Fallback: WEBRAIN_USER / WEBRAIN_PASS env vars.
List what’s in the vault (names only, never secrets):

​
Persistent profiles

webrain launch <service> <profile> [url] spawns a stealth Chrome with a persistent per-account profile, keyed by "service:profile" so webrain_login re-attaches. webrain_close_launch kills it (profile + cookies remain for the next launch).

​
Save / restore auth state

Export cookies + localStorage to a state file so a login follows you across machines:
  1. Log in in real Chrome (webrain_login / webrain launch).
  2. Export: webrain_cookies (includes HttpOnly) or webrain cookies --port 9222 --out file.
  3. Import: webrain_setcookies on the session connection.
  4. webrain_batch without cdp_urls — set + batch must share one connection (obscura isolates cookie contexts per CDP connection).
Session cookies die on Chrome restart — export on the live authenticated browser before closing it.