Repository navigation
Shobr
Job hunting means repeating the same search, triage and resume tailoring across many postings. shobr is a Python command-line tool (MIT) that automates LinkedIn job search with a human in the loop. It discovers postings, enriches their details, scores them, tailors resumes and tracks pipeline status, and it requires explicit approval before submitting any application.
The outcome is less manual triage with the final decision kept by the user. To avoid looking like automation, shobr drives the user's authenticated daily browser through the beachpatrol tool instead of a headless browser. It needs Python 3.14, the uv package manager, groff and pandoc, and depends on any-llm-sdk and roffume for language-model access and resume compilation.
The project states two risks itself. Automating a logged-in daily browser may violate LinkedIn's terms, and resume data is sent to language-model providers. The repository was created on 2026-09-27 and has 14 stars, one open issue and about 3,500 lines of code with end-to-end tests.
It overlaps with JobNavigator in purpose, and relies on browser automation of the kind covered by BrowserUse.
shobr sits in Assess with an outer position. It shows a careful human-approval design, but it is very new, has little adoption and carries terms-of-service exposure, and there is no first-person use.
