SPT Documentation

Appearance

SPTFlo Documentation ​

Enterprise AI agent orchestration for the SPT ecosystem.

Overview ​

SPTFlo provides 60+ specialized AI agents coordinated through swarms with self-learning hooks, vector memory, fault-tolerant consensus, and MCP integration. It is the orchestration layer of the SPT platform, consuming SPT's core vector database and neural architecture components.

Documentation Structure ​

Architecture ​

Package Documentation ​

Individual module documentation:

  • CLI - Main CLI with 39 commands and 27 MCP tools
  • Shared - Common types, events, utilities
  • MCP - Model Context Protocol server
  • Memory - SptDB + SQLite hybrid memory
  • Embeddings - Vector embeddings and semantic search
  • Swarm - 15-agent hierarchical mesh coordination
  • Neural - SONA self-optimizing neural architecture
  • Hooks - Self-learning lifecycle hooks
  • Performance - Benchmarking and Flash Attention
  • Guidance - Guidance Control Plane
  • Security - Security scanning and CVE detection
  • AI Defence - AI manipulation defense
  • Browser - Browser automation
  • Codex - Code intelligence engine
  • Plugins - Plugin SDK and management
  • Testing - TDD framework
  • Deployment - Release management
  • Claims - Collaborative issue claims
  • Integration - Integration patterns
  • Providers - Model provider implementations

Plugin Documentation ​

Domain-specific plugins:

Guides ​

Reference ​

Examples ​

  • V2 Examples - 25+ example projects including REST APIs, chat apps, data pipelines, ML foundations, and parallel processing demos

Released under the MIT License.

Copyright 2024-present SmartPoints Technology