OpenApply is a local-first application: it runs entirely on your own
machine, binds its web server to 127.0.0.1 only, and stores your personal
data (resume, contact info, work authorization answers, application history)
under user_data/ on your own disk - never on a remote server the project
controls, because there isn't one.
That said, it does three security-sensitive things worth being careful about: it parses files you upload (resume parsing), it drives a real browser against real third-party websites (application automation), and it can optionally call third-party AI APIs (Anthropic/OpenAI/a local endpoint) with your API key.
If you find a security issue - especially anything involving:
- path traversal or arbitrary file write via the resume upload endpoint
- credentials, API keys, or session cookies leaking into logs or the
user_data/files that agit add .might accidentally pick up - the FastAPI server binding to anything other than
127.0.0.1by default - the Playwright application engine submitting a form without the user-review step, or bypassing/solving a CAPTCHA
please open a private report via GitHub's "Report a vulnerability" feature on the repository's Security tab, rather than a public issue. Include steps to reproduce and the affected version. We'll acknowledge reports within a few days.
Only the latest released version is supported with security fixes while the project is pre-1.0.
These are treated as security/ethics boundaries, not missing features - see
README.md's Privacy & Security section for the full list:
- No CAPTCHA/bot-check bypassing.
- No automatic submission on LinkedIn or Indeed, ever.
- No fabricating citizenship, work authorization, sponsorship, clearance, or other sensitive answers on your behalf.
- No sending your data anywhere by default - AI providers are opt-in and only ever see job-description text and your own confirmed profile data when you explicitly enable a provider.