Repolode
Submit
RikyZ90

RikyZ90/ShibaClaw

Self-hosted personal AI agent with a built-in WebUI, memory, tools, MCP support, sub-agents and scheduled automations, using local or cloud LLMs.

Open-source alternative to:ChatGPT

View on GitHubWebsite
Stars
85
Forks
12
Open issues
5
Last commit
3 days ago

What is it?

ShibaClaw is a personal AI assistant you install and run on your own machine or server. It bundles a web interface, a Windows desktop launcher and a mobile-friendly layout, so you can chat with the agent from a browser on your LAN.

The agent can use Ollama, LM Studio, llama.cpp, vLLM or any OpenAI-compatible endpoint, as well as cloud providers such as OpenAI, Anthropic, DeepSeek and OpenRouter. It keeps editable long-term memory, session history and reusable skills, and can search an optional FAISS-based document knowledge base. File, shell and web tools are included, and external tools can be added through the Model Context Protocol.

For longer work, it supports isolated or forked sub-agents, one-shot, interval and cron schedules, and messaging channels such as Discord, Telegram, Slack and Matrix. Security features include an encrypted credential vault, workspace-restricted file access, SSRF filtering and prompt-injection defenses, which the project notes reduce but do not eliminate risk.

Key features

  • 01Built-in WebUI with a Windows desktop launcher and mobile layout
  • 02Supports local models via Ollama or OpenAI-compatible servers, plus cloud providers
  • 03Editable long-term memory, session history and reusable skills
  • 04Optional document knowledge base with FAISS search
  • 05MCP client for connecting external tools
  • 06One-shot, interval and cron automations with sub-agents

Use cases

  • Running a private AI assistant on a home server with a local model
  • Automating recurring tasks such as scheduled checks or reports
  • Querying your own PDFs and Markdown notes through a knowledge base
  • Connecting an AI agent to Discord, Telegram or Slack for team or personal messaging

Quick start

  1. 1

    Install the package with pip (requires Python 3.12+)

    pip install shibaclaw
  2. 2

    Start the WebUI and agent engine, then open http://localhost:3000

    shibaclaw web --with-gateway

Things to know

  • Requires Python 3.12+ for the pip route or Docker for the container route.
  • Some account connections are unofficial or restricted, such as the Google Gemini CLI integration and Claude subscription OAuth.
  • Using a cloud model or remote MCP service sends relevant content to that external provider.

In collections

Open-source AI agent frameworks & appsMCP serversSelf-hosted appsRun LLMs locallyOpen-source personal AI agents

Topics

#ai-agent#self-hosted#llm#mcp#ollama#personal-assistant#rag#automation#agent-framework#ai-agents

FAQ

What is ShibaClaw?
Self-hosted personal AI agent with a built-in WebUI, memory, tools, MCP support, sub-agents and scheduled automations, using local or cloud LLMs. ShibaClaw is a personal AI assistant you install and run on your own machine or server. It bundles a web interface, a Windows desktop launcher and a mobile-friendly layout, so you can chat with the agent from a browser on your LAN.
Is ShibaClaw free to use? What license does it have?
ShibaClaw is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is ShibaClaw still maintained?
Actively maintained. The last commit to ShibaClaw was on Nov 3, 2026.
What can I use ShibaClaw for?
Running a private AI assistant on a home server with a local model Automating recurring tasks such as scheduled checks or reports Querying your own PDFs and Markdown notes through a knowledge base Connecting an AI agent to Discord, Telegram or Slack for team or personal messaging
How do I install ShibaClaw?
Install the package with pip (requires Python 3.12+): pip install shibaclaw Start the WebUI and agent engine, then open http://localhost:3000: shibaclaw web --with-gateway
Where can I get ShibaClaw?
The source code and installation instructions are on GitHub: https://github.com/RikyZ90/ShibaClaw

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects