Skip to content

Argus Intelligence Platform v1.1 - USB Portable Edition

Latest
Latest

Choose a tag to compare

View all tags
@Taimaishu Taimaishu released this 04 Jan 04:20
· 3 commits to main since this release
v1.1
7be3acb

Argus Intelligence Platform v1.1 - USB Portable Edition

Release Date: January 3, 2026

🚀 What's New

USB Portable Deployment

  • Complete portable package ready for USB drives or external storage
  • Cross-platform support - Works on Windows 10/11 and Linux (Ubuntu, Debian, Fedora, Arch)
  • Automatic setup wizard guides users through first-run configuration
  • No installation required - Just extract and run!
  • 903 KB package containing 206 files with complete application

Safe Improvements to Image/Entity Search

  • 100% behavior preservation - All existing functionality unchanged
  • Enhanced stability - Defensive guards prevent crashes from invalid inputs
  • Better logging - Query transformations and name mappings now visible
  • Image validation - MIME type checking, size limits (10 MB max), HEAD request validation
  • Database safety - Proper transaction rollback handling with SQLAlchemyError
  • Debug metadata - Enhanced observability for troubleshooting

New Features

  • Setup Wizard - First-run experience for API key configuration
  • Settings Page - Manage API keys for OpenAI, Anthropic, Google, Unsplash, Pexels
  • Canvas Generation - AI-powered entity clustering and timeline visualization
  • Knowledge Graph - Relationship mapping across documents
  • Metadata Analysis - Cross-document entity tracking and co-occurrence analysis
  • Entity Extraction - Enhanced NER with support for 9+ entity types
  • Unredaction Service - Document recovery and redaction removal
  • Canvas Chat - Entity-specific conversations with context awareness
  • Entity Nodes - Specialized UI for Person, Organization, Location, Event, etc.

📦 Download

USB Portable Package: argus-usb-portable.zip (903 KB)

Quick Start

Windows:

  1. Extract argus-usb-portable.zip to your USB drive
  2. Double-click START_ARGUS_USB.bat
  3. Follow the on-screen prompts
  4. Browser opens automatically

Linux:

  1. Extract argus-usb-portable.zip to your USB drive
  2. Open terminal in the extracted folder
  3. Run: bash START_ARGUS_USB.sh
  4. Follow the on-screen prompts
  5. Browser opens automatically

✨ Key Improvements

Image Search Pipeline

  • Added defensive guards for null/empty query strings
  • Enhanced logging shows query transformations (e.g., "epstein" → "Jeffrey Epstein financier")
  • Image validation with MIME type checking (JPEG, PNG, GIF, WebP, SVG)
  • Size limit enforcement (10 MB maximum)
  • HEAD request validation before downloading images
  • Better timeout handling with explicit exception catching

Entity Enrichment

  • Input validation prevents crashes from invalid entity names or node IDs
  • Database transaction safety with proper rollback on errors
  • Enhanced logging for name enhancement (e.g., "Andrew" → "Prince Andrew Duke of York")
  • Defensive guards for missing documents in evidence collection
  • Better error context in all exception handlers

Preserved Critical Behaviors

All intentional behaviors remain unchanged:

  • ✅ Name mappings (epstein, clinton, andrew, trump, maxwell)
  • ✅ Query enhancement keywords (mugshot, official portrait, etc.)
  • ✅ 70% Wikipedia validation threshold
  • ✅ Source priority order (wikipedia → google → pexels → unsplash)
  • ✅ Skip types (date, event, phone, email, address)
  • ✅ Regex patterns for entity name extraction
  • ✅ AI theory generation threshold (5 mentions minimum)
  • ✅ Character limits for database fields

🔧 Technical Details

Testing

  • ✅ 13/13 API tests passed (100% pass rate)
  • ✅ Zero behavioral changes detected
  • ✅ All existing functionality preserved
  • ✅ Original files backed up (.backup files)

New Services

  • canvas_generation_service.py - AI-powered canvas generation
  • knowledge_graph_service.py - Relationship mapping
  • metadata_analysis_service.py - Cross-document tracking
  • entity_extraction_service.py - Enhanced NER
  • entity_enrichment_service.py - Photo and metadata enrichment (improved)
  • image_search_service.py - Multi-provider image search (improved)
  • unredaction_service.py - Document recovery

Frontend Components

  • SetupWizard.tsx - First-run configuration
  • SettingsPage.tsx - API key management
  • CanvasChatPanel.tsx - Entity-specific chat
  • EntityDetailPanel.tsx - Enriched entity information
  • Entity nodes: Person, Organization, Location, Event, Date, Email, Phone, Address, Vehicle, Financial

Documentation

  • IMAGE_SEARCH_PIPELINE_REVIEW.md - Complete safety review (13 KB)
  • INTENTIONAL_BEHAVIOR_DO_NOT_CHANGE.md - Protection guide (12 KB)
  • DISTRIBUTION_README.md - USB deployment overview
  • CREATE_USB_DEPLOYMENT.md - Detailed deployment instructions
  • USB_DEPLOYMENT_CHECKLIST.md - Pre-deployment verification
  • LAUNCHER_GUIDE.md - Launcher script documentation

📋 System Requirements

Minimum

  • OS: Windows 10/11 or Linux (any modern distro)
  • Python: 3.10 or higher
  • RAM: 2 GB
  • Storage: 500 MB (more for documents)
  • Internet: Required for initial setup and API-based AI

Recommended

  • RAM: 8 GB (for large document analysis)
  • Storage: 5 GB+ (for Ollama + document storage)
  • CPU: Multi-core for faster processing

🔑 API Keys

The setup wizard supports:

  • OpenAI - Best performance, $0.50/million tokens
  • Anthropic Claude - High quality, good for analysis
  • Google AI - Gemini models support
  • Ollama - Free, offline, privacy-focused (no API key needed!)

🛠️ Migration Guide

If upgrading from v1.0:

  1. Backup your existing installation
  2. Download argus-usb-portable.zip
  3. Extract to new location (or replace existing)
  4. Your data is stored separately (database, uploads)
  5. No migration required - just run the new version!

Note: If you have an existing .env file, it will be preserved.

🐛 Bug Fixes

  • Fixed crash when image search receives null/empty queries
  • Fixed database rollback issues in entity enrichment
  • Fixed timeout handling in image fetch operations
  • Fixed missing document errors in evidence extraction
  • Improved error messages throughout the application

📚 Resources

🙏 Credits

  • Built with FastAPI, React, TypeScript, Python
  • AI powered by OpenAI, Anthropic, Google, and Ollama
  • Icons by Lucide React
  • PDF parsing with PyMuPDF
  • Entity extraction with spaCy

📝 Full Changelog

Added:

  • USB portable deployment package (903 KB, 206 files)
  • Setup wizard for first-run API key configuration
  • Settings page for API key management
  • Canvas generation service with AI clustering
  • Knowledge graph service for relationship mapping
  • Metadata analysis service for cross-document tracking
  • Entity extraction service with enhanced NER
  • Entity enrichment service with automatic photo/metadata
  • Image search service with multi-provider support
  • Unredaction service for document recovery
  • Canvas chat panel for entity-specific conversations
  • Entity detail panel with enriched information
  • Specialized entity nodes (9+ types)
  • Database migrations system
  • Comprehensive documentation (5+ new guides)

Improved:

  • Image search stability with defensive guards
  • Entity enrichment error handling
  • Database transaction safety
  • Logging for query transformations
  • Input validation across all services
  • Timeout handling for external requests
  • Error messages and context

Fixed:

  • Crash from null/empty queries
  • Database rollback failures
  • Missing document errors
  • Timeout exceptions in image fetch

Preserved:

  • 100% of existing image/entity search behavior
  • All critical name mappings and query enhancements
  • Wikipedia validation threshold (70%)
  • Source priority ordering
  • Entity classification logic
  • AI theory generation thresholds

Made with ❤️ for investigators, researchers, and analysts

🤖 Generated with Claude Code
Co-Authored-By: Claude Sonnet 4.5 noreply@anthropic.com

Assets 3
Loading