Waku Agent: Legible, Local-First AI Agent Harness in Python

RepositoryShenSeanChen/waku-agentTool

Waku Agent is a local-first AI assistant harness designed to avoid bloated framework abstractions. Built in straightforward Python, it implements an agent loop, tiered memory in SQLite, and built-in evaluations in an easily readable codebase.

Key points

  1. Four architectural pillars: harness, execution loop (~95 lines of Python), tiered memory, and eval/LLM-Ops.
  2. Local-first storage: Semantic, episodic, and procedural memory reside in a single inspectable SQLite file.
  3. Cockpit & CLI: Includes terminal access and a browser-based dashboard running at localhost:7777.
  4. Integrated evaluation: Offers deterministic tests alongside LLM-as-a-judge workflows and release gates.
  5. Features extensions such as a Telegram gateway and wake-word support.

Why it matters

Many popular agent frameworks hide fundamental mechanics behind heavy abstractions. Waku demonstrates how to implement a capable, controllable personal assistant with minimal code that developers can read and adapt in an afternoon.

Analysis

The project addresses framework fatigue in the AI agent space by stripping down agent logic to essentials without sacrificing key capabilities. It is aimed at engineers seeking an understandable foundation for personal agents or practical study of agent loops and memory gating. Built with Python, SQLite, and uv, it integrates with providers like OpenAI and Anthropic. With over 1,600 GitHub stars, PyPI distribution, and an active release flow, Waku presents a mature and focused harness rather than an sprawling enterprise platform.

Mentioned links

Found in

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Video

    Video:Sean‘s AI Stories

    Loop vs Graph Engineering for AI Agents: Concepts and Practice with Waku Agent

    Sean Chen addresses the debate surrounding loop versus graph engineering in AI agent systems. Presenting an engineering ladder from prompts to graphs, he explains when to apply each pattern and demonstrates the architecture using the open-source Waku Agent project.

    AI & LLMs· Tutorial

  • Repository:Significant-Gravitas/AutoGPT

    AutoGPT: Open-Source Platform for Building and Running Autonomous AI Agents

    AutoGPT is a widely known open-source project designed to build, deploy, and run autonomous AI agents. Users can define end-to-end workflows via natural language prompts ('AutoPilot') or a visual builder, executing tasks on demand, on a schedule, or via triggers.

    187.1kstarsPython

    AI & LLMs· Tool

  • Repository:NousResearch/hermes-agent

    Hermes Agent: Self-Improving Open Source AI Agent by Nous Research

    Hermes Agent is a cross-platform, self-improving AI agent designed with an autonomous learning loop, cross-session memory, and integration into multiple messaging platforms and terminal interfaces.

    240.8kstarsPython

    AI & LLMs· Tool

  • Repository:AgriciDaniel/claude-obsidian

    claude-obsidian: Local-First AI Knowledge Management for Obsidian and Claude Code

    claude-obsidian is an MIT-licensed open-source project that connects Claude Code with Obsidian into a self-organizing knowledge graph. Implementing Karpathy's LLM Wiki pattern, it maintains a structured vault made entirely of local plain Markdown files.

    14.6kstarsPython

    AI & LLMs· Tool

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.