★彡[MDAA - Update Log]彡★

v1.5.6 : New features powered by AI & New System Prompt File 19 November 2025

NEW Refine prompt with PES 🤔: This feature makes it more convenient for users to use it if they can't think of a prompt. They can use it to improve it and make it more detailed.
NEW New System Prompt File 🤔: Improved the personality file system from json to toon to minimize tokens for users.
NEW Supports Google's latest model now. 🤔: Gemini 3.0 Pro Preview

NEW Deep Think Mode 🤔: Activated via @deepthink, this mode specializes in abstract reasoning, hypothesis generation, and 'outside-the-box' thinking for complex security scenarios.
NEW Deep Research Mode 🔎: Activated via @deepresearch, this mode functions as a Tier-3 Senior Threat Analyst, performing deep technical deconstruction of artifacts, logs, or code with mandatory MITRE ATT&CK mapping and IOC generation.
NEW Learn with Guidance Mode 🎓: Activated via @learnwithguidance, this mode acts as an AI tutor, breaking down complex topics step-by-step and checking for understanding before proceeding. Ideal for learning new concepts or analyzing specific file content with guidance.
IMPROVED Integration Popup & Mode Switching ✨: Typing '@' now smoothly presents available modes (Deep Think, Deep Research, Learn with Guidance) and integrations. Selecting a mode switches the AI's persona for the current chat session.

NEW Predictive Threat Modeling 🤖: Proactively simulate attacks against your known defenses (Playbooks) to find gaps before attackers do.
FIXED Fixed bug in callGeminiAPI function 🔗: Fixed an issue and improved performance for connecting this function to the file upload feature.

NEW Support for larger language models 🤖: Added Gemini 2.5 Flash Lite , Gemini 2.0 Flash , Gemini 2.0 Flash Lite
NEW Added file upload feature 🧠: You can upload files for Agnet to analyze your files.
NEW Sub-features after MDAA answer 🤖: Added Copy Text , Copy Source , Read Aloud
NEW New Platform 🧠: Knowledge Base - A knowledge base that can be written down for Agents to refer to, such as official documentation or any textual information that provides additional knowledge for Agents.
NEW New Feature Chat 🧠: Add Microphone & Increase interaction with Agents
IMPROVED Update Settings Page 🌐: Supports a total of 6 spoken languages for microphone
IMPROVED Update Threat Hunting Hub 🌐: Threat Actor Playbooks - Adding a list of attackers , CVE Exploit Watchlist - Adding vulnerabilities to watch for
IMPROVED Improve MDAA Status 🌐: Add clarifications

NEW AbuseIPDB Daily Blacklist 🔑: Added AbuseIPDB API Key feature and added Key setting on settings page.
NEW Anti-Account Takeover (ATO) 🛡️: Implemented a security lockdown. If a session unlock fails, the user is restricted to the Settings page, forcing a secure key reset.
IMPROVED Password Policy & UI 🔑: Enforced a strong password complexity policy and added a real-time strength meter to guide users.
IMPROVED Web App Hardening 🌐: Added a strict Content Security Policy (CSP) and Subresource Integrity (SRI) to all external scripts, preventing XSS and supply-chain attacks.
IMPROVED Fixed logic in IoC Collector 🔗: Improved IoC Collector detection logic
FIXED Security Bypass 🐛: Patched a vulnerability where the security lockdown could be bypassed by using the "Forgot Password" feature.
FIXED Correct the faulty logic 🔗: We found an error with the My Case Link Sharing feature.

FIXED Collaboration Hub 🤝: Added an Export Case button that works the same way in Chat.
FIXED CORS Origin Error 🔗: Fixed a limited CORS integration feature issue.

NEW Collaboration Hub 🧑‍🤝‍🧑: Launch of collaborative workspaces through file and link sharing with a Case Management Dashboard, Team Workflow Board, and Shared Intelligence Center.

NEW AI-Generated Analyst Notebook ⚡️: Added a "Generate Notes" button that uses AI to summarize chat sessions into the Analyst Notebook based on selected playbooks.
IMPROVED Smart Integration Popup ✨: Enhanced the chat input to show a smooth, animated popup list of available integrations when the user types `@`.
NEW Direct VirusTotal Integration & Live Threat Feed 📡: Rolled out direct VT API checks from the Workbench and a live cybersecurity news feed on the dashboard.

NEW Core Platforms 🛠️: Launched the Threat Hunting Hub, Analysis Workbench, and Training Ground to create a comprehensive analysis ecosystem.
IMPROVED Dashboard Overhaul 📈: Added the "MDAA Status" section to monitor version, AI model, latency, and real-time server status.
IMPROVED AI Model Selection 🧠: Added an option to switch between Gemini 2.5 Flash and Gemini 2.5 Pro models in Settings.

NEW Contextual Actions 👆: Added action buttons below AI responses like "Generate Rule," "Suggest Mitigations," and "Extract IoCs" for a faster workflow.
NEW TTP Fingerprinting & CVE Correlation 🕵️: The AI can now identify potential threat actors from TTPs and correlate behavior with relevant CVEs.

NEW Core Conversational AI 🤖: Launched MDAA as an AI assistant for malware analysis.
NEW Client-Side Architecture 💻: Designed to run entirely in the user's browser for maximum security and privacy.
NEW Secure API Key Storage 🔒: Implemented Web Crypto API (AES-GCM) to securely encrypt and store API keys locally.