🤖 Yet Another OpenClaw Clone (YAOCC)

YAOCC is a lightweight, Go-based agentic AI assistant inspired by OpenClaw.

It is designed to be sandboxed, Docker-friendly, and has a tiny footprint (4-5mb RAM, 10mb binary), running smoothly even on ARMv7 devices.

Features

• LLM Integration: Connects to Ollama, OpenRouter, or any OpenAI-compatible provider.
• Web Search: Support for SearxNG, Brave, and Perplexity with fallback mechanisms.
• Skill System: Dynamic CLI command execution based on user requests. It can even create its own skills!
• Persistent Memory: Maintains conversation history via session files. Sessions can be summarized to reduce context.
• Telegram Support: Integrated bot with long-polling.
• Swagger UI: API documentation available at /docs.

Disclaimer: 90% of this code was made by vibecoding using Google's Antigravity.

Quick Start

Prerequisites

• Go 1.23 or higher (only for building the app)
• Ollama (running locally) or valid API keys for other providers.

Usage

1. Initialize Project:

   ./yaocc init

   This generates necessary workspace files and a default config.json. make sure to set your config before using the app. the config file can use expanded variables coming from .env file.

2. Start Server:

   ./yaocc-server

3. Chat:

   ./yaocc chat "Hello, who are you?"

please refer to DOCS.md. to know more about how to enable Telegram

Installation

Build files can be found in the Releases tab.

To build manually:

# Build CLI
go build -o yaocc.exe ./cmd/yaocc

# Build Server
go build -o yaocc-server.exe ./cmd/yaocc-server

Docker

Building

You can build the image using Docker Compose:

docker compose build

Or manually using docker build with the default arguments:

docker build \
  --build-arg YAOCC_BASE_IMAGE="node:24-alpine" \
  --build-arg YAOCC_DOCKER_APK_PACKAGES="git nano curl unzip python3" \
  --build-arg YAOCC_DOCKER_RUN_COMMANDS="" \
  --build-arg YAOCC_USER="node" \
  -t yaocc:latest .

Using Docker Compose (Recommended)

Simply run:

docker compose up -d

Using Docker Directly

You can run the container directly, ensuring you mount your data directory and set the configuration path:

docker run -d \
  --name yaocc \
  -e YAOCC_CONFIG_DIR=/app/data \
  -v ./data:/app/data \
  -p 8080:8080 \
  yaocc:latest

Documentation

For detailed configuration options, API usage, Docker deployment, and advanced features, please refer to DOCS.md.

License

This project is licensed under the MIT License - see the LICENSE file for details.