5 items
Prepare LinkedIn invite drafts, plus optional AI interview prep and company research using your own Anthropic or OpenAI API key. - Career Connect is an extension for the part of networking that normally becomes a tedious collection of tabs, copy-and-paste, and half-finished notes. It helps you: - find relevant LinkedIn people with precise search criteria; - prepare personalized connection invites at scale; - research a company and role before an interview; and - turn a public professional profile, your CV, and a job description into practical interview preparation. The result is a focused workflow from discover → understand → reach out. It keeps the human in control of the final message and the final decision. Career preparation with your own context The popup opens on a Career tab (with Connect — the connection workflow — alongside it); the last tab you used is restored the next time you open the popup. Career Tools use a bring-your-own-key model: the user chooses Anthropic or OpenAI and pays the provider directly. The extension provides: - one combined, adaptive interview-preparation report built from whatever context is available — CV, job description, interviewer profile, company details, role, seniority, location, and interview stage — rather than separate interview/company reports; - a constrained, searchable model selector per provider (type to filter; only listed, verified model IDs can be selected — no free-text model IDs); - a company-research stage when a valid LinkedIn company URL is available, and a synthesis stage that still produces a full report without it; - best-effort extraction from the current page for job, interviewer/profile, and company context — including LinkedIn profiles (including your own), job postings, job-search result pages, company pages, and other LinkedIn/non-LinkedIn pages with useful information — that only ever adds to existing data, never overwrites it; - streamed reports in a dedicated report view, including a collapsed "Generation context" section showing the exact input snapshot and provider/model used; - report/case history: reopen a prior report; add its saved context to your current case or replace the current case with it; delete one you no longer need; or start a New case without losing earlier ones — History also shows total storage used and flags any unreadable saved record; a session-storage recovery anchor for the rare case a report can't be saved to local storage right away — it stays recoverable (with a Retry save action in History and on the report page) until the browser closes; - source-aware output and explicit distinctions between researched facts, user-provided context, and modeled estimates; provider-specific request handling, error messages, and separate saved credentials per provider; and - a transmission preview and an explicit transmission notice before any content leaves the browser — reviewed once per submission, with no separate recurring consent checkbox. Career Tools are optional and the core connection workflow works without an API key.
rating_count is the Chrome Web Store ratings count, not a written-review count.