🦀 HyperClaw Cookbook
GitHubGitHub (opens in a new tab)
  • Introduction
    • Start Here
    • 1 — Identity
    • 2 — ZenBin
    • 3 — Web Search
    • 4 — Memory
    • 5 — Scheduling
    • 6 — Morning Brief
    • 7 — Clawvisor
    • Quiet Mode
    • Troubleshooting
    • Overview
    • Installing Skills
    • /goal — The Gold Skill
    • Agent Browser
    • Humanizer
    • Self-Improving Agent
    • OpenAI Whisper
    • Nano Banana Pro
    • Nano PDF
    • Grill Me
    • Write Your Own Skill
    • Overview
    • Connect Your Integrations (Clawvisor)
    • Agent Browser (Browserbase)
    • Give Your Claw an Email Address
    • Power Model (GPT-5.5)
    • Content Creation (Slides & Video)
    • Voice
    • Indexable Memory (QMD)
    • ZenBin — Publish & Message
    • Self-Learning Agent
    • Overview
    • Developer Agent
    • SRE Agent
    • Graphic Designer Agent
    • Documentation Agent
    • Household Agent
    • Overview
    • Quiet Mode & Clean Output
    • Manage Claws via API
    • Prompting as a Skill (25 Prompts)
    • Overview
    • HyperClaw, OpenClaw & the Rise of the Agent
    • Models
    • Agents
    • Agent Harnesses
    • Memory & Context
    • Tools & Actions
    • History
    • Overview
      • Course Overview
        • Module 1: Reverse Prompting
        • Why Your First Prompt Is Usually Wrong
        • The Conversation Mindset
        • Ask the Agent to Ask You Questions
        • Context Before Commands
        • Iteration: Refining vs. Starting Over
        • Reading the Agent's Response
        • Breaking Big Requests into Conversations
        • Building Your Prompting Instincts
        • Module 2: Memory & Context
        • The Blank Slate Problem
        • What Memory Actually Is
        • Writing Your First Memory Note
        • What Is a Skill?
        • Writing Your First Skill
        • Personal Context vs. Task Context
        • Building a Skill Library Over Time
        • When Memory and Skills Change Everything
        • Module 3: Multi-Agent & Delegation
        • Why One Agent Can't Do Everything Well
        • Thinking in Roles
        • Your First Delegation
        • Parallel Work vs. Sequential Work
        • How Agents Hand Off to Each Other
        • When to Supervise vs. When to Trust
        • The Future You're Preparing For
    • Course Overview
      • Module 1: Reverse Prompting
      • Why Your First Prompt Is Usually Wrong
      • The Conversation Mindset
      • Ask the Agent to Ask You Questions
      • Context Before Commands
      • Iteration: Refining vs. Starting Over
      • Reading the Agent's Response
      • Breaking Big Requests into Conversations
      • Building Your Prompting Instincts
      • Module 2: Memory & Context
      • The Blank Slate Problem
      • What Memory Actually Is
      • Writing Your First Memory Note
      • What Is a Skill?
      • Writing Your First Skill
      • Personal Context vs. Task Context
      • Building a Skill Library Over Time
      • When Memory and Skills Change Everything
      • Module 3: Multi-Agent & Delegation
      • Why One Agent Can't Do Everything Well
      • Thinking in Roles
      • Your First Delegation
      • Parallel Work vs. Sequential Work
      • How Agents Hand Off to Each Other
      • When to Supervise vs. When to Trust
      • The Future You're Preparing For
Edit this page on GitHub →
💡 Tips & Patterns
Overview

Tips & Patterns

Hard-won patterns from daily Claw use. These aren't step-by-step recipes — they're the kind of advice that changes how you work with your agent.

Quiet Mode & Clean Output→Manage Claws via API→
Household AgentQuiet Mode & Clean Output
HyperClaw Cookbook — built with Nextra · Powered by HyperClaw