How rxmna8502/vybe-intelligence-vault works
Vybe Intelligence Vault harvests AI-related repositories and research, enriches and evaluates candidate material, publishes curated Markdown nodes, then builds a searchable semantic graph. Engineers and agents consume that knowledge through the MCP gateway; the 3D explorer provides a visual route. README documentation describes these product capabilities, while sampled source confirms crawler enrichment, index construction and state persistence. MCP and world implementation details were not sampled, so their internal wiring is shown only at the documented subsystem level.
Main entry points: evaluate_repo.py Discovery and Evaluation · build-index.js Index and Graph Builder · vault-core Vault Index · state-manager.js Vault State Manager
Components 18
Harvest and Evaluate
- Discovery and Evaluation
scripts/evaluate_repo.py - Web Enrichment
scripts/crawler.py
Knowledge Graph
- Index and Graph Builder
scripts/build-index.js - Vault State Manager
scripts/state-manager.js - Vault Index
vault-core - Embedding Cache
Knowledge Access
- MCP Gateway
mcp-server/mcp_server - 3D Graph Explorer
world - Published Indexes
_index
Knowledge Content
- Agent Resources
ai/agents - RAG Resources
ai/rag - Other Taxonomies
ai - Daily Digests
daily-digests
Other components
- AI Engineer
- AI Agent
- Research Sources
- GitHub
- LLM Evaluators
Connections 21
- Research Sources → Discovery and Evaluation: supplies sources
- GitHub → Discovery and Evaluation: repository discovery
- Discovery and Evaluation → Web Enrichment: enriches URLs
- Discovery and Evaluation → LLM Evaluators: requests evaluation
- LLM Evaluators → Discovery and Evaluation: returns evaluation
- Discovery and Evaluation → Agent Resources: publishes nodes
- Discovery and Evaluation → RAG Resources: publishes nodes
- Discovery and Evaluation → Other Taxonomies: publishes nodes
- Discovery and Evaluation → Daily Digests: generates digests
- Index and Graph Builder → Agent Resources: reads Markdown
- Index and Graph Builder → RAG Resources: reads Markdown
- Index and Graph Builder → Other Taxonomies: reads Markdown
- Index and Graph Builder → Daily Digests: reads Markdown
- Index and Graph Builder → Embedding Cache: caches vectors
- Index and Graph Builder → Vault State Manager: loads and stores state
- Vault State Manager → Vault Index: reads and writes
- AI Agent → MCP Gateway: queries context
- AI Engineer → 3D Graph Explorer: explores graph
- MCP Gateway → Vault Index: queries index
- 3D Graph Explorer → Vault Index: visualizes graph
- Index and Graph Builder → Published Indexes: publishes indexes