Login profiles

Isolated Chrome profile dirs for the platforms that need a login — for scraping, for Easy Apply auto-submit, or both. Round-robin rotation across profiles per run; one flag doesn't sink the others.

Scrape + apply (login powers both): linkedin · naukri · wellfound · hirist · instahyre. The linkedin profile additionally powers /linkedin-human warm-intro scraping AND future LinkedIn Easy Apply automation; Wellfound shares it per §8.166.

Apply-only (scrape runs anonymously, login reserved for Easy Apply auto-submit): indeed · glassdoor · ziprecruiter. Pre-log-in now so when apply automation ships, the profiles are already warm.

Public-anonymous scrapers (linkedin_public, indeed scrape, glassdoor scrape, ziprecruiter scrape, ats_direct) need no profile.

Add a login profile

Creates an empty per-platform chrome_profile dir. You log in manually afterward — credentials never leave your browser.

Registered burners

How rotation works

  • Each new run picks the active burner with the oldest last_used for that platform.
  • Switching burners per cell mid-session triggers anti-bot signals (fingerprint changes mid-flight). We rotate per-RUN only.
  • If an account starts misbehaving, mark it flagged here and the picker will skip it.
  • Set CV_AGENT_RUN_BURNER=<name> to pin a specific burner for one run (overrides rotation).
Loading mobile proxy…