Media
1 item
Description
Enterprise-grade job application verification engine — fraud-resistant detection with HMAC proof 1. Active Session Job Details Recovery On confirmation pages, standard HTML job title selectors are usually absent, causing fallback to generic titles like "Thank you for applying" or "Candidate Home". Real-Time Extraction: Extracts the real jobTitle and company from the initial application form page. Active Session Sync: Saves details inside the active session context in local storage. Auto-Recovery on Redirect: Dynamically recovers the real job details from the active tab session when evaluating the confirmation page, falling back to a clean "Job Application" only if still generic. 2. Reference ID DOM Sanitization DOM Tag Stripping: Excludes <script>, <style>, <noscript>, <svg>, <link>, and <iframe> contents from textContent prior to reference regex extraction. Technical Code Filtering: Explicitly ignores internal code placeholders (like TRACKINGPIXELHTML or TEMPLATE) to ensure only actual human-facing reference codes are parsed. 3. Clean Verification History Cache Verified-Only Storage: Restricts local storage verification history (verifications_v2) to successfully verified entries. Auto-Purge: Automatically deletes cached entries if a re-check evaluates them as unverified or if scores drop below the threshold, preventing UI clutter. 4. Dynamic Environment Key Integration Vite .env integration: Supports VITE_EXTENSION_API_KEY inside extension/.env and resolves it dynamically at compile time via import.meta.env in EngineConfig.ts. 5. Backend Auth Integration & Self-Healing Event Sync Bearer Token Resolution: Resolves and parses incoming JWT Bearer tokens to log Chrome Extension sessions under the correct recruiter's name. Active User Fallback: Queries and matches the first active database administrator/recruiter as a fallback to satisfy foreign key constraints. Placeholder Session Auto-Creation: Dynamically creates parent VerificationSession rows if event uploads (addEvents / finalizeSession) are received for a session that doesn't exist yet, eliminating database constraint errors.
rating_count is the Chrome Web Store ratings count, not a written-review count.