47 lines
1.7 KiB
Markdown
47 lines
1.7 KiB
Markdown
# LLM-Optimized Databank
|
|
|
|
This directory contains all databank information formatted for optimal LLM consumption. Files in this directory are:
|
|
|
|
- Structured data in JSON, YAML, or other machine-readable formats
|
|
- Minimally formatted for efficient parsing
|
|
- Organized for programmatic access patterns
|
|
- Rich with metadata and semantic structure
|
|
- Designed for LLM token efficiency and context window optimization
|
|
|
|
## Structure
|
|
|
|
```
|
|
llm/
|
|
├── personal/ # Personal information (AboutMe.json, TSYS.yaml, etc.)
|
|
├── agents/ # AI agent guidelines and tools (structured)
|
|
├── context/ # General context information (structured)
|
|
├── operations/ # Operational environment information (structured)
|
|
├── templates/ # Template files (structured)
|
|
├── coo/ # Chief Operating Officer information (structured)
|
|
├── cto/ # Chief Technology Officer information (structured)
|
|
└── README.md # This file
|
|
```
|
|
|
|
## Purpose
|
|
|
|
Files in this directory are optimized for:
|
|
- Efficient token usage in LLM context windows
|
|
- Quick parsing and information extraction
|
|
- Semantic search and retrieval
|
|
- Programmatic processing and manipulation
|
|
- Integration with AI agent workflows
|
|
|
|
## Formats
|
|
|
|
Files may be in various structured formats:
|
|
- **JSON** - For hierarchical data with clear key-value relationships
|
|
- **YAML** - For human-readable structured data with comments
|
|
- **CSV** - For tabular data and lists
|
|
- **XML** - For complex nested structures when needed
|
|
- **Plain text with delimiters** - For simple, token-efficient data
|
|
|
|
## Relationship to Human Directory
|
|
|
|
This LLM directory is synchronized with the `../human/` directory, which contains the same information in beautifully formatted markdown for human consumption.
|
|
|
|
--- |