TnsAI Docs
Official documentation for the TnsAI framework
Architecture | Features | Methodology
Overview
This repository contains the official documentation for the TnsAI Framework — a comprehensive multi-agent system built on BDI architecture and Gaia methodology for Java 21+.
Documentation Structure
Architecture
| Document | Description |
|---|---|
| LAM Pattern | Language-Action Model design pattern |
Features
| Document | Description |
|---|---|
| Core | Agents, Roles, Tools, Streaming, Events, Security, Knowledge, Prompts, Output Parsing, Variants, Resilience, Memory, Errors |
| LLM | Providers, Routing, Caching, Cost Tracking, Audio |
| Tools | Tool Catalog, Custom Tools |
| Coordination | Topologies, Workflows, Council & Voting, Negotiation, Protocols, Judge Agent, Advanced |
| Intelligence | FSM, Planning, Reasoning, Context & Sessions, Learning, RAG, AutoTeamBuilder |
| Quality | Observability, Security, Validation |
| MCP | Client, Server, Transports, Registry |
| Server | WebSocket, RAG Pipeline, Tool Approval |
| Integration | SCOP |
| Evaluation | Benchmarking, G-Eval, agentic evaluators, quality gates, auto-harness loop |
| Semantic Annotations | 55+ annotation system for declarative agent design |
Methodology
| Document | Description |
|---|---|
| TnsAI Methodology | BDI + Gaia methodology guide |
Architecture (Internal)
| Document | Description |
|---|---|
| CI/CD Rules | Build and deployment pipeline rules |
| Feature-Module Matrix | Feature to module mapping |
| Module Audit | Module health and audit report |
| Module Map | Module dependency and responsibility map |
Research
| Document | Description |
|---|---|
| Competitor Analysis | Comparison with other agent frameworks |
| Research Reports | Research findings and technical notes |
Roadmap
Framework Modules
| Module | Repository | Description |
|---|---|---|
| Core | TnsAI.Core | Agent, Role, Action, LLM abstractions |
| LLM | TnsAI.LLM | 14 LLM provider implementations |
| Coordination | TnsAI.Coordination | Multi-agent orchestration |
| Intelligence | TnsAI.Intelligence | FSM, Planning, Reasoning |
| Quality | TnsAI.Quality | Observability, Security, Validation |
| Tools | TnsAI.Tools | 130+ built-in tools |
| MCP | TnsAI.MCP | Model Context Protocol |
| Integration | TnsAI.Integration | Integration bridges and tests |
| Server | TnsAI.Server | Javalin backend, hybrid RAG, WebSocket |
| Evaluation | TnsAI.Evaluation | Agent evaluation and benchmarking |
| Web | TnsAI.Web | Documentation website (Next.js) |
| Docs | TnsAI.Docs | Official documentation |
| Papers | TnsAI.Papers | Research publications |
Contributing
Documentation contributions are welcome. Please follow these guidelines:
- All documentation must be written in English
- Use clear, concise language
- Include code examples where applicable
- Follow the existing document structure
License
Part of the TnsAI Framework — Building Intelligent Multi-Agent Systems