OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Frameworks

Waku Agent

Waku Agent is an MIT-licensed, Python-based local-first AI agent harness for running a personal assistant with a readable agent loop, SQLite-backed semantic, episodic and procedural memory, dashboards and evaluation components.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Waku Agent — OpenViST project profile in Agent Frameworks.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Waku Agent when the goal is to study, modify or self-host an agent harness rather than only consume a finished assistant product. Its local-first SQLite design is useful when inspectable memory and a compact loop are central requirements.

Who it is for

  • Developers building a self-hosted personal assistant.
  • People learning or prototyping agent-loop and memory designs.
  • Teams inspecting assistant behavior with deterministic tests or LLM-as-judge evaluation.
  • Users who want an agent codebase they can read and modify.

Documented uses

  • Run a personal assistant from a terminal or local browser dashboard.
  • Store assistant memory in an inspectable SQLite file.
  • Separate semantic, episodic and procedural memory records.
  • Trace and evaluate assistant turns with deterministic and LLM-based methods.
  • Experiment with provider adapters and readable agent architecture.
Before you choose

Requirements & limits

The repository is public, MIT-licensed and provides package-install and clone-based setup instructions. The documented workflow requires a provider key; connecting Waku Memory adds a browser sign-in step.

  • The loop, memory structure and evaluation components are exposed as code, making the project suitable for inspection and modification.
  • Local memory is stored in one SQLite file, while Waku Memory is presented as a separate hosted shared-memory connection.
  • The framework supports terminal and browser-dashboard entry points, with additional gateways listed in the repository.
  • Provider setup is environment-based and the repository lists several provider options, but the supplied material does not specify exact model interfaces.
  • Replacing SQLite with remote or encrypted storage is Requires modification because only the local SQLite design is documented.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install and start Waku Agent locally

Set up the documented terminal and dashboard entry points for a local Waku Agent instance.

Before you start

  • Python package installation with pip, or a cloned repository with uv available.
  • One provider key configured when prompted or through .env.
  • A Linux, macOS or Windows environment with command-line access.

Steps

  1. Install the package with pip, or clone the repository and install it in editable mode with uv.

  2. If using the repository setup, copy .env.example to .env, choose a provider and add one key.

  3. Run the waku command for terminal interaction.

  4. Run waku dashboard to open the browser cockpit at localhost:7777.

Expected outcome

A locally running Waku Agent accessible through the terminal or its browser dashboard.

Guide limitations

  • The supplied instructions do not specify exact model interfaces or provider-specific compatibility.
  • The guide documents setup and entry points, not a verification of successful execution.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
ShenSeanChen/waku-agent · Official homepage title · retained in its source language
How we review sourcesSuggest a correction