LLMs.txt - Complete Documentation for AI Models¶

The LLMs.txt file contains the complete framework documentation optimized for Large Language Model consumption.

Overview¶

LLMs.txt is a comprehensive, structured documentation file designed specifically for AI models to understand and use the ReAct Agent Framework effectively.

File Location: LLMs.txt

Size: 800+ lines Format: Plain text with clear structure Coverage: All versions from v0.1.0 to v0.12.0

What's Included¶

1. Overview & Quick Start¶

Complete introduction with working examples

2. Installation¶

All installation methods and options

3. Core Concepts¶

ReAct pattern, agent creation, tool registration

4. Memory Systems (v0.10.0+)¶

ChatMemory (SimpleChatMemory, SQLiteChatMemory)
KnowledgeMemory (ChromaKnowledgeMemory, FAISSKnowledgeMemory)
Complete API and examples

5. Multi-Provider Support¶

OpenAI, Anthropic, Google, Ollama configuration

6. Built-in Tools¶

Search, filesystem, computation tools

7. Objectives System¶

Goal tracking and management

8. Reasoning Strategies¶

ReAct, ReWOO, Reflection, Plan-Execute

9. Environments¶

Web, CLI, File environment interaction

10. MCP Integration¶

Model Context Protocol server setup

11. Layer 4: Infrastructure (v0.11.0)¶

Complete production infrastructure: - Monitoring: Metrics, logging, telemetry - Resilience: Retry, circuit breaker, fallback, timeout - Security: RBAC, sandbox, audit, secrets - Cost Control: Budget, rate limiting, quotas - Human-in-the-Loop: Approval, intervention, feedback

12. Layer 3: Multi-Agent Systems (v0.12.0)¶

Complete multi-agent capabilities: - Communication: Messages, protocols, channels - Orchestration: Orchestrator, workflows, task delegation - Coordination: Shared state, consensus, locking - Collaboration: Teams, patterns, negotiation

13. Complete API Reference¶

All classes, methods, parameters, and return types

14. Examples & Use Cases¶

5 complete, production-ready examples: 1. Simple Q&A Agent 2. Research Agent with Memory 3. Production Agent with Infrastructure 4. Multi-Agent System 5. Complete Enterprise System

Why LLMs.txt?¶

For AI Models¶

Structured format optimized for AI consumption
Complete coverage of all features in one file
Working examples with full context
No navigation required - everything in sequence
Version tracking - clear indication of feature versions

For Developers¶

Quick reference for AI pair programming
Copy-paste ready code examples
Architecture overview in one place
Version history and migration paths

How to Use¶

As a Developer¶

1. Share with AI assistants:

# Copy to clipboard
cat LLMs.txt | pbcopy  # macOS
cat LLMs.txt | xclip   # Linux

2. Reference in prompts:

"Using the ReAct Agent Framework as described in LLMs.txt,
create an agent that..."

3. Quick lookup:

# Find specific topics
grep -A 10 "Multi-Agent" LLMs.txt
grep -A 20 "Memory Systems" LLMs.txt

As an AI Model¶

The file is structured with clear section markers:

================================================================================
SECTION NAME
================================================================================

Each section includes: - Concept explanation - Code examples - API signatures - Best practices - Common patterns

Example Content Structure¶

================================================================================
4. MEMORY SYSTEMS (v0.10.0+)
================================================================================

IMPORTANT: v0.10.0 separated memory into TWO types:
- Chat Memory: Sequential conversation history
- Knowledge Memory: Semantic knowledge base (RAG)

CHAT MEMORY:
------------

SimpleChatMemory (Development):
```python
from react_agent_framework.core.memory.chat import SimpleChatMemory

agent = ReactAgent(
    name="Agent",
    chat_memory=SimpleChatMemory(max_messages=100)
)

[... complete examples and API ...]

---

## Keeping Updated

The LLMs.txt file is updated with each release:

- **v0.12.0**: Added Layer 3 (Multi-Agent Systems)
- **v0.11.0**: Added Layer 4 (Infrastructure)
- **v0.10.0**: Added Memory System Refactoring
- **v0.9.0**: Initial LLMs.txt creation

---

## Download

**Direct link**: [LLMs.txt on GitHub](https://github.com/marcosf63/react-agent-framework/blob/main/LLMs.txt)

**Raw file**: [Raw LLMs.txt](https://raw.githubusercontent.com/marcosf63/react-agent-framework/main/LLMs.txt)

```bash
# Download with curl
curl -O https://raw.githubusercontent.com/marcosf63/react-agent-framework/main/LLMs.txt

# Download with wget
wget https://raw.githubusercontent.com/marcosf63/react-agent-framework/main/LLMs.txt

Contribute¶

Found an issue or want to improve LLMs.txt?

The file is located at the repository root: /LLMs.txt