Tsukuyomi: Intelligent Modular Framework for Structured Analysis and Processing

Intelligent Modular Framework for Structured Analysis and Processing

TSUKUYOMI is an advanced modular intelligence framework designed for systematic analysis, processing, and reporting across multiple domains. Built on a component-based architecture, it separates operational logic from presentation through specialized modules and personality cores, enabling flexible and extensible analytical workflows.

• Modular Architecture: Self-contained operational modules for specific analysis tasks
• Personality System: Customizable communication styles and interaction patterns
• Structured Output: Standardized formatting with confidence levels and source attribution
• Domain Expertise: Specialized modules for OSINT, economic analysis, infrastructure assessment, and strategic intelligence
• Flexible Workflows: Modules can be chained and combined for complex analytical processes

Technical Documentation

Claude

Note: Paid Subscription is more or less required

• Project Knowledge Methods
  • Adding Files Manually, simple & effective - lacks 'Sync' functionality
  • Adding GitHub Repository (Best Method) - allows usage whilst keeping updated modules, allows hotswapping modules mid-session.
• Individual Chat Method (Janky)
  • If lacking subscription, this is the only method that will work: you will have to manually add the files to the chat via the upload.
  • Can be time extensive with free 'limits'.
  • Only allows 4 Sonnet, Opus performs better.

Gemini

Note: Paid Subscription makes life easier, but isn't entirely necessary, I am not a Gemini user but I have learned this to help others navigate this.

• Project Knowledge Methods
  • Manual File Method, ditto from Claude
  • Cloning the GitHub repository to a Drive folder, and loading it this way. This is the advised method, I lack the One Plus subscription but I imagine this could be automated fairly easily to 'Sync'.

ChatGPT

Note: Paid Subscription more or less mandatory. (Important: GPT seems to be lagging behind in support for this kind of thing, or at least that's what I can see as a 'free' user.)

• Project Knowledge Methods
  • Cloning the GitHub repository to a Drive folder, this should provide basic Sync functionality like Gemini.

Others

Note: 100% speculation here, aided by Claude 4.

//AMATERASU: Enterprise-focused with extensive Microsoft integration.

How it works:

• You can upload your own documents for use as a knowledge source at the agent level, which your agent then uses to generate answers with generative AI
• Copilot Studio: Create custom agents with knowledge bases
• Microsoft 365 Integration: Access files from SharePoint, OneDrive
• Enterprise knowledge management across Microsoft ecosystem

File Support: Files larger than 512 MB aren't supported. Files with encryption, such as sensitivity labels or password protection, aren't supported Requirements: Microsoft 365 Copilot license or Copilot Studio access

Best for: Organizations heavily invested in Microsoft ecosystem

//AMATERASU: Research-focused with real-time web integration.

How it works:

• Pro users can leverage advanced models like Claude 3 and GPT-4 to handle longer files
• Upload academic papers to extract key insights, summarize findings, and explore related topics
• Combines uploaded documents with real-time web search
• 30-day file retention

File Support: PDFs, CSV files, text documents, academic papers Requirements: Perplexity Pro ($20/month) for unlimited file uploads; limited free access

Best for: Research-intensive workflows requiring real-time information

//AMATERASU: European AI solution with strong document capabilities.

How it works:

• Upload, organize, and analyze arbitrary files using powerful OCR and vision models under the hood
• Mistral OCR is an Optical Character Recognition API that sets a new standard in document understanding
• Advanced document processing and OCR capabilities
• Enterprise knowledge base integration

File Support: PDFs, images, documents with OCR capabilities Requirements: Various pricing tiers available

Best for: Document-heavy workflows requiring advanced OCR and European data residency

1. Claude Projects - Upload all .tsukuyomi files directly
2. Enter: "Initialise Amaterasu" or provide your data
3. Framework auto-activates with full functionality

• ChatGPT Projects: Upload files, create project-specific instructions
• Gemini Gems: Create custom Gem with TSUKUYOMI files as knowledge base
• Copilot Studio: Build custom agent with TSUKUYOMI modules as knowledge sources
• Perplexity Pro: Upload for research-focused intelligence analysis

• File Limits: Vary by platform (Claude: context window, Gemini: 10 files, etc.)
• Persistence: Most platforms retain project knowledge across sessions
• Cost: Free tiers available but paid plans recommended for full functionality
• Integration: Consider your existing tool ecosystem

• Core System (tsukuyomi_core.tsukuyomi): Central orchestration component managing module loading, execution, and transitions
• Personality Cores (*_personalitycore.tsukuyomi): Define communication style, tone, and user interaction patterns
• Operational Modules (*.tsukuyomi): Contain specific analysis, processing, or reporting functionality
• Activation Key (key.activationkey): Framework initialization and file interpretation instructions

The AMATERASU personality core provides:

• Analytical Focus: Strong pattern recognition and evidence-based reasoning
• Precision: Values accuracy and specificity, avoiding ambiguity
• Professional Courtesy: Maintains appropriate formality in interactions
• Efficiency: Prioritizes clarity and relevance
• Adaptability: Adjusts communication style based on context and user expertise

• Data Recognition & Ingestion: File type identification and entity extraction
• Discipline Alignment: OSINT methodology classification
• Functional Inference: Analytical task recommendation
• Correlation Analysis: High-confidence relationship identification
• Output Summarization: Findings synthesis and next steps

• E1: Economic Vulnerability Assessment: Systematic evaluation of economic weaknesses and dependencies
• E2: Trade Network Impact Analysis: Assessment of trade relationship changes and disruptions
• E3: Resource Security Analysis: Critical resource dependencies and supply chain vulnerabilities
• E4: Financial Stability Assessment: Financial system risks and contagion analysis

• S1: Strategic Scenario Modeling: Evidence-based geopolitical outcome scenarios
• S2: Actor Capability Assessment: Systematic evaluation of geopolitical actor capabilities
• S3: Strategic Impact Projection: Consequence analysis across multiple domains
• S4: Multi-factor Trend Analysis: Interlocking geopolitical trend identification

• Infrastructure Assessment: Comprehensive evaluation of infrastructure systems
• Utility Monitoring: Systematic monitoring of utility services
• Infrastructure Dependency Mapping: Relationship analysis between infrastructure systems
• Critical Infrastructure Vulnerability: Security assessment of critical infrastructure

• Flight Data Analysis: Aviation tracking data processing and interpretation
• Web Search & OSINT Collection: Systematic open-source intelligence gathering
• Comprehensive Reporting: Professional-grade intelligence report synthesis
• Micro-Summary Authoring: Concise 280-character intelligence summaries
• PDF Document Generator: Professional document creation with customizable layouts

The framework initializes automatically when the activation key is processed:

1. Loads tsukuyomi_core.tsukuyomi
2. Initializes the AMATERASU personality core
3. Discovers and loads all available modules
4. Presents greeting and awaits user input

Modules follow a standardized execution sequence with structured outputs:

1. Data Ingestion: Upload or provide data for analysis
2. Discipline Alignment: Framework determines appropriate OSINT methodology
3. Functional Inference: Recommends optimal analytical operations
4. Specialized Analysis: Executes domain-specific modules as needed
5. Correlation & Synthesis: Identifies relationships and patterns
6. Reporting: Generates structured output or comprehensive reports

All framework components use the .tsukuyomi extension with JSON-like structure:

• Structured Tags: All outputs use standardized prefixes for clarity
• Confidence Levels: Analytical assessments include reliability indicators
• Source Attribution: Information properly cited with confidence ratings
• Ethical Constraints: No reproduction of copyrighted content beyond fair use

TSUKUYOMI follows core principles of:

• Modularity: Each component is self-contained but interoperable
• Explicit Structure: Clear operational sequences and expected outputs
• Separation of Concerns: Operational logic separated from presentation
• Analytical Rigor: Evidence-based reasoning with appropriate confidence qualifiers
• Ethical Intelligence: Responsible information handling and attribution

1. Create a new .tsukuyomi file following the standard schema
2. Define execution sequence with specific actions and outputs
3. Specify dependencies and personality compatibility
4. Test module integration with existing framework

1. Create a new personality core with _personalitycore.tsukuyomi suffix
2. Define communication style, terminology preferences, and formatting rules
3. Specify adaptive parameters for different contexts
4. Integrate with core personality management system

This project is licensed under the MIT License - see the LICENSE file for details.

Contributions are welcome! Please read our contributing guidelines and code of conduct before submitting pull requests.

For support, documentation, or questions about the TSUKUYOMI framework, please open an issue in this repository.

TSUKUYOMI: Bringing structured intelligence to complex analytical challenges