Robort Gabriel - AI Expert

Roadmap to master OpenClaw.

Learn the stack in order: run a Gateway, shape agent workspaces, add skills and plugins safely, wire channels, automate with Lobster, then harden for production.

Getting started with OpenClaw.

From zero to a working agent in seven small steps: install the Gateway, create a workspace, configure it, send your first message, add a skill, connect a channel, then ship.

Comment START to get step by step guide.

Follow and comment " all" to get 90+ detailed commands.

Comment CLAUDE to get full step by step guide for your business.

Most people use Claude Cowork out of the box. That’s why it feels like a "smart book" instead of a "smart worker." To get actual results, you need to set up your Global Instructions and a claude.md file.

My Setup Ritual:
✅ Global Instructions: I tell Claude exactly who I am (Senior AI Architect) so I never have to repeat my role or brand voice.
✅ The claude.md File: This is the brain of your folder. Put your brand guidelines and project goals here so Claude hits the ground running.
✅ Chunk Your Skills: Separate your tasks. One for Writing, one for Sales, one for DevOps.

Stop fighting your AI and start orchestrating it.
Share and follow for more Claude tips.

Comment GUIDE to get my full setup.

It’s weekend, let’s create one openclaw ai agent. 😎
Comment one below!

I built an OpenClaw AI agent that monitors Reddit, drafts replies, and waits for your approval before posting.

What it does:

· Monitors subreddits you care about (DevOps, cybersecurity, content marketing, etc.)
· Sends a Telegram digest of new posts
· Generates two draft comment options per post
· You approve one (or none) – never posts automatically
· Posts via browser after your approval

Why HITL (human-in-the-loop)?
Because auto-posting on Reddit is risky. This keeps you in control.

Built for: Experts, consultants, founders who want to build reputation on Reddit without the daily grind.

👇 Comment REDDIT to get the GitHub link free.

What’s one thing hate about Openclaw ?

Claude just got a MASSIVE compute upgrade (partnering with SpaceX).

Effective today:

1️⃣ Claude Code; Doubled 5-hour rate limits (Pro, Max, Team plans)
2️⃣ Peak hours limit reduction: REMOVED for Pro & Max plans
3️⃣ API rate limits: Substantially raised for Opus models

Why this matters for OpenClaw builders:

I build OpenClaw agents that run on Claude (via OpenRouter). These changes mean:

✅ Longer agent runs: No more hitting limits mid-pipeline
✅ Peak hour reliability: Agents run anytime, not just off-hours
✅ More Opus usage. Better reasoning for competitor spy, content, and data agents

Real impact example:

My Competitor Spy agent used to monitor 10 pages before hitting limits. Now? 25-30 pages easy.

How to make your OpenClaw AI agent sound exactly like you (and never sound like a generic robot again)

This is one of the smartest moves you can make with OpenClaw.

Step-by-step:
1. Open a new Google Doc (or Notion page)
2. Paste the full WRITING RULES prompt below
3. Save it as YOURNAME-Writing-Style.md (or anti-ai-writing-style.md)
4. Upload this .md file into your OpenClaw memory / skills folder
5. Tell your agent: “From now on, always follow the rules in my Writing Style file”

Your OpenClaw will now write posts, emails, replies, content — everything — in your real voice.

# WRITING RULES
Read this before writing to me or for me.
Goal: write with context, taste, and a reason to speak.
Apply with judgment. Spirit over letter. Clean natural writing wins.
0. Rule priority
Use this order when rules collide:
1. Be accurate.
2. Be clear.
3. Be specific.
4. Sound human.
5. Use style only when it improves the sentence.
Do not follow a style rule so strictly that the result gets awkward.

1. Default voice
Write directly, specifically, and naturally.
Start with the useful answer.
Use short paragraphs. 1 or 2 sentences by default.
Vary rhythm. Short sentence. Longer sentence. Fragments allowed when natural.
Use contractions: don’t, can’t, won’t, it’s, you’re.
Use I and you when natural. Talk to people.
Prefer active voice.
Be specific. Use numbers, names, concrete details, dates, places, prices, etc.
Use plain uncertainty when unsure (I think, probably, maybe…).
If the point is made, stop.

Pro Tip for OpenClaw users:
Combine this with your USER.md and SOUL.md files. This turns your agent from “smart” to “actually sounds like me.”

Comment OPENCLAW and I’ll send you:
• The full ready-to-use prompt (clean version)
• My recommended folder structure for OpenClaw
• Bonus: How to create your own personal voice file fast

Real talk…

What’s one thing in your business that is quietly leaking money every month?

Common answers I get:
• Paying staff to do tasks AI can handle
• Losing sales because of slow response time
• Wasting hours creating content instead of closing clients
• Manual invoicing & payment chasing
• Expensive tools you barely use

Comment your biggest money leak below 👇

Comment OPENCLAUDE if you want me to send you the exact AI agent setups my clients are using to plug these leaks.