Waku Agent

Waku Agent is an open-source personal assistant codebase that runs locally and manages agent loops, memory, and evaluation logic for developers.

Waku Agent provides a locally executable framework for personal assistants, focusing on transparency by allowing users to read the entire logic in Python. It features three memory types—semantic, episodic, and procedural—stored in a single SQLite file that users can access directly. The system monitors agent thought processes via a local dashboard, visualizing messages as they move through the logic flow. It includes built-in evaluation tools, using deterministic tests and LLM-as-judge logic to verify performance before release. Developers can integrate various models, including Anthropic, OpenAI, Gemini, DeepSeek, and others, by updating a single provider variable.

Waku Agent at a glance

Best for: Software engineer
Runs on: CLI, open source
Pricing: Open source; MIT license.
Released: July 2026
Category: AI agent tool
Website: github.com/ShenSeanChen/waku-agent

Key features

Local-first SQLite memory storage
Python-based logic loop
Semantic, episodic, and procedural memory support
Integrated deterministic and LLM-as-judge evaluation
Dashboard for monitoring agent thought processes
Gateway support for terminal, voice, Telegram, Discord, and WhatsApp

Use it for

Developing custom AI agent memory systems
Testing agent reasoning with deterministic evals
Managing AI agent workflows on local hardware
Building cross-platform agent integrations

Alternatives to Waku Agent

Beeuni — Agents retain context and improve skills over time through a controlled review process.
MyAgent — It features built-in permission controls, long-term memory, and skill self-evolution.
Alook — Connects existing local coding agents to team communication channels.
memU — It gives every connected agent a shared and inspectable memory layer.
Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
Rakazo — Complete ownership of data, models, and execution environment.
Agent Squad — Runs multi-agent orchestration in Python, TypeScript, and locally on Apple devices.
Apache Maka — It provides a complete, append-only record of every agent action and decision.
Agent Beacon — Retains coding agent knowledge across tools so errors do not repeat.
Tigerless Agent Memory — Maintains agent history in portable, greppable Markdown files without API keys.
Atomic Agents — Build predictable, modular AI applications with consistent, reliable outputs.
OmniAgent — It evolves skills in real-time and dynamically hardens security during task execution.

All Waku Agent alternatives