aibuddy Documentation
Start here
Run your first task
Install, configure, and start one agent task first, so you can confirm aibuddy works in your environment.
LLM + context + tools = brain, eyes, hands — build the mental model from the formula and the ReAct loop.
Follow one task through runtime, context, memory, tools, multi-agent coordination, delivery, and operations.
Framework-agnostic theory and industry practice — the ETCLOVG survey, and Anthropic's and AWS's engineering experience.
The Agent System
A theory-first learning path distilled from the reference book — what an agent is, context, memory, tools, evaluation, continuous evolution, and multi-agent collaboration.
- 01.1 Agent Basics
- 01.2 Context Engineering
- 01.3 Memory & Knowledge
- 01.4 Tools
- 01.5 Evaluation
- 01.6 Continuous Evolution
- 01.7 Multi-Agent
How aibuddy turns agent capabilities into a working product — runtime, context, memory, tools, multi-agent coordination, delivery, platform boundaries, and operations.
- 02.1 Quick Start
- 02.2 System Backbone
- 02.3 Agent
- 02.4 Auth & Identity
- 02.5 Operations
- 02.6 Implementation Reference
Universals
Framework-agnostic research and theory, aibuddy-free — the ETCLOVG harness survey, Skills design, model-side foundations, and Anthropic's engineering practice.
- 01.1 ETCLOVG
- 01.2 Model-Side Foundations
- 01.3 Agent Skills
- 01.4 Anthropic Engineering
How an agent product turns task execution into a business: cost, metrics, pricing, growth, operations, and compliance.