Documentation
Your privacy is our No.1 priority.
The complete guide to the AegisVectro Security Architecture v3.3.
AegisVectro is a comprehensive security suite that runs directly inside your browser. By combining Local Heuristics (speed) with Generative AI (intelligence), we offer protection against threats that static blocklists typically miss. We have recently upgraded to a massive 17-Engine Defense Matrix powered by Google Gemini models to protect your digital footprint.
Flagship Edition: Privacy First
Architecture
Unlike cloud-based antivirus solutions that send your browsing history to a remote server, the AegisVectro extension runs entirely within your Chrome browser. It intercepts and analyzes the DOM (Document Object Model) locally.
BYOK (Bring Your Own Key) AI
To enable the advanced Generative AI features (Engines 1, 2, 7, 9, 10, 11), users must provide their own Gemini API Key from Google AI Studio. This guarantees:
- Zero Middlemen: Your data goes directly from your browser to Google's API. AegisVectro Labs never sees or touches it.
- Privacy Compliant: By default, Google does not use data submitted via API keys to train their models.
- Context-Aware AI: If enabled in settings, you can share real-time telemetry (tracker stats, blocked popups) to provide the AI with context without sacrificing your privacy to a centralized server.
Settings & Core Protection
Engine Sensitivity
You can fine-tune how strict the AI agent is when analyzing websites:
- Smart (Recommended): Uses high AI intelligence to differentiate normal e-commerce features (like standard marketing) from genuinely manipulative threats.
- Literal (Strict): Extremely strict mode that flags any deviation from ideal security practices without giving the benefit of the doubt. Useful for highly sensitive browsing.
Share Context with AI
This setting sends background tracker and advertisement statistics to your chat agent for a significantly better conversational experience. You can turn this on or off at any time.
For example: You can ask "How many ads did you block here?" and the AI will answer accurately using the extension's live Radar data.
Core Protection Options
AegisVectro provides multiple layers of automated defense that you can toggle on or off in the settings menu:
- Auto-Scan Pages: Automatically initiates a deep threat analysis the moment a website loads.
- Auto-Reject Cookies: Automatically hunts for and actively clicks 'Reject All' or 'Decline' on GDPR/CCPA consent banners, protecting your privacy from tracking cookies.
- Block Ad Networks: Intercepts and completely blocks intrusive advertisements and known ad-servers at the network level, speeding up your browsing.
- Block Popups & Redirects: Prevents malicious websites from hijacking your browser window, stopping spam popups and forced invisible redirects.
The 17-Engine Defense Matrix
1. Visual Guard AI
Powered by Google Gemini models, this engine acts like a human eye. While code-level scanners can be tricked by obfuscation, Visual Guard literally "looks" at a screenshot of the page. It detects if a site visually mimics a known brand (like PayPal or Microsoft) but is hosted on a deceptive domain.
2. Ask Aegis (Context-Aware Chat)
An integrated AI assistant powered by Google Gemini models. You can ask questions about the current page, like "Is there a refund policy?" or "Is this a secure checkout?". As mentioned above, it can securely use your local telemetry context to answer data-specific questions.
4. Tracker Radar
A sophisticated network interceptor utilizing the Manifest V3 declarativeNetRequest API.
It blocks over 20+ major analytics and advertising trackers (like Google Analytics, Facebook Pixel,
and DoubleClick) before they even load.
5. Breach Checker
Allows you to query the current domain against known major data breaches. Instantly know if the service you are signing up for has a history of poor data security.
6. Link Hover Shield
Before you click, AegisVectro analyzes the link. It displays a real-time tooltip if the link points to a dangerous executable (.exe, .bat), a URL shortener (bit.ly), or an unencrypted HTTP destination.
7. Legitimacy Check
Uses Google Gemini models to analyze the intent and business logic of the website. It cross-references the page content to verify if it acts like a legitimate business entity or a fly-by-night scam operation.
8. DOM Watchdog
A background MutationObserver that monitors your active tabs. It detects hidden iframes
dynamically injected by malware and alerts you if background scripts attempt to secretly modify your
password or credit card fields without your focus.
9. Privacy AI
Sends the website's text to Google Gemini models to specifically hunt for abusive privacy policies. It looks for clauses that allow the site to sell your data to third parties without consent.
10. Dark Patterns
Identifies deceptive UI practices designed to trick users. This includes fake countdown timers, hidden subscription fees, and "roach motel" designs where cancellation is deliberately difficult.
11. Sentiment Scan
Analyzes the linguistic tone of the webpage. Highly urgent language ("Your PC is infected! Call now!") is strongly correlated with tech support scams and phishing.
12. Malware Guard
A robust local heuristic engine that scans all downloadable links on the page. It cross-references file extensions against a matrix of trusted software domains to block unauthorized executable downloads.
13. Network Ad Blocker
A serverless ad-blocking layer running on Manifest V3. It intercepts requests to known ad networks (PopAds, PropellerAds) ensuring a cleaner, faster browsing experience.
14. Pop-up Guard
Hooks into the browser's window.open method to neutralize aggressive, unprompted popup
windows and redirect loops often used by spam sites.
15. URL Analysis
Locally analyzes the domain structure. Detects raw IP addresses, excessive subdomains, risky TLDs (.tk, .ml), and Punycode homograph attacks (xn--).
16. Image OCR
When used with the Vision AI, this engine pulls embedded text out of images. Scammers often put text inside images to bypass text-based security filters; this engine defeats that tactic.
17. Spam Detection
Analyzes the density of keywords, excessive capitalization, and hidden input fields. Pages heavily overloaded with hidden tracking fields or spam keywords are flagged instantly.
Try the Interface
Interact with the simulation below to see how AegisVectro presents this data to you.
Legal & Compliance
Privacy Policies
We believe in full transparency regarding how your data is handled. Please read our Privacy Policy to understand how AegisVectro maintains zero-data exfiltration while utilizing the BYOK AI framework.