waku-agent
What it is
Waku Agent is a local-first AI agent harness that provides a legible, code-based implementation of an agent loop, memory system, and evaluation framework. It runs entirely on the user's machine using SQLite for memory and supports multiple LLM providers via a simple adapter layer.
The problem it solves
Most AI agent frameworks are opaque, cloud-dependent, or complex, making it difficult for users to understand, debug, or truly own their agent's logic and data.
What you could use it for
- Building a personal AI assistant with transparent, inspectable logic
- Learning agent architecture by reading a minimal, legible codebase
- Developing local-first applications with persistent semantic and episodic memory
- Running local LLM evaluations with deterministic and LLM-as-judge tests
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.