The Hiring Panel
Four adversarial AI reviewers stress-test your resume against a target role, then hand back concrete, scored feedback. Generate the prompt, run it in your agent, ingest the verdict.
DiagnoserRecruiterRewriterHiring Manager
Open-source job-search copilot
Your agent job-hunts while you sleep.
ApplyOS is a self-hosted copilot. Your own Claude agent finds and scores roles on a schedule, and you approve every move. You wake up to a ranked shortlist.
Self-host in one command
$ docker compose up -d- Runs on your machine, against your own Claude subscription.
- No per-seat fees and no third-party API metering.
- Every automated action is logged and reversible.
One loop, run for you every morning.
- 1
Hunt
Your agent searches on a schedule and pushes new roles into your queue.
- 2
Score
Deterministic fit scoring ranks every role against your candidate profile.
- 3
Review
You triage the shortlist. Nothing reaches an application without your approval.
- 4
Stress-test
The Hiring Panel runs four adversarial reviewers over your resume.
A real pipeline, not another job board.
Deterministic fit scoring
Explainable 0 to 100 scores per role, with a per-signal breakdown. No black-box ranking.
Review queue + scheduler
A human-in-the-loop workflow and agent tasks that run on a cadence you set.
Yours to host
One docker compose brings up the database, backend, and agent runner. Your data and your Claude key never leave your box.
Every run is audited
No blind auto-apply. Each agent action is logged with its input, output, and timing.
Self-host it this weekend.
Create an account to get started, or star the repo and self-host the whole stack yourself.