# Agentic AI Architect

A production-tested knowledge base for deploying OpenClaw with Claude Code, covering server deployment, messaging-app integration, model configuration and security hardening, plus example AI agent business workflows.

Agentic AI Architect is a knowledge base built from production deployments of OpenClaw combined with Claude Code. It documents server deployment with one-command and manual setup scripts, remote access through Tailscale, gateway authentication, security hardening checklists and troubleshooting guides, and connecting agents to messaging channels such as WeChat, WeCom and Feishu.

It also documents model configuration for cost control, including routing through domestic model providers and coding-plan subscriptions, designing an agent's persona through a layered file, configuring heartbeat and scheduled tasks, and writing custom skill files.

The repository includes a set of example agent business workflows built on this stack: automated social-media publishing, financial invoice recognition and reporting, a multi-platform morning-briefing assistant, meeting-note capture for a CRM, a quantitative-research pipeline, and a GitHub secret-leak auditor. It is written in Python.

Website: https://github.com/DjangoPeng/agentic-ai