Operate persistent agents without handing them the steering wheel.
OpenClaw is Turtleand's applied lab for public-safe notes on scheduling, routing, voice loops, sub-agents, and the boundaries that keep automation useful.
Learning path
Choose a field note, then follow the system underneath.
Module 0: Setup & Safety
Getting started with persistent AI agents safely
Module 0: Setup & Safety
Getting started with persistent AI agents safely
Claude CLI Integration
Run Claude CLI commands through OpenClaw and monitor your Anthropic usage limits.
Cron Jobs & Scheduled Automation
Schedule recurring tasks, reminders, and automated check-ins with OpenClaw's built-in cron system.
Delegating Development to Your Agent
Configure Git access so OpenClaw can develop on your behalf while you're away.
Persistent AI agents
Deploy an always-on AI agent on a cloud server with messaging and security.
Sub-Agents & Session Isolation
How sub-agents work, why they're isolated from your main session, and what happens when you change settings mid-flight.
Fix Voice Message Speed Control in OpenClaw + Telegram
Convert TTS output to OGG Opus so Telegram voice messages get proper speed control buttons.
Voice notes with Telegram
Send voice messages to your AI agent and have them transcribed and executed.
Module 1: Foundations
Core patterns for working with AI agents
Module 1: Foundations
Core patterns for working with AI agents
AI Image Gen vs OpenClaw + Code
Same banner brief, two approaches: a standard image model vs OpenClaw generating code. The results tell you where AI actually adds value.
Cost tracking for AI agent operations
How to monitor and optimize spending when your AI agent runs on pay-per-token APIs.
Migrating cron jobs to a new model provider
How to audit, classify, and migrate scheduled AI tasks when switching providers.
Why your AI agent should never depend on one provider
What happens when your only model provider disappears, and how to design for it.
Setting up multi-model routing
How to use different AI models for different tasks instead of running everything on one.
Sub-agent reliability: timeouts, monitoring, and recovery
What happens when a sub-agent hangs, and the three-layer system to handle it.