Every OpenClaw Use Case I Could Find (85+)

I went through 148 replies to Lenny's tweet, the Clawverse gallery, and Brandon Wang's writeup. Here's every use case I found.

Some of these are practical. Some are unhinged. One person gave their agent $1,000 and told it to start a business. Another built a meme battle arena where the agent fought itself for an entire night.

I studied engineering in school but spent my whole career as a founder, growth person, and PM. I've been running OpenClaw for about a month now with an agent named Alfred on my MacBook Air. I'm nowhere near most of these people. But scrolling through what everyone's building gave me a pretty clear picture of where this is all going.

So here's the list. 85+ use cases, organized by category, with attribution to the people who shared them.

The Big Themes

Four patterns kept showing up across almost every reply:

Always-on agents. Most power users run OpenClaw 24/7 on a Mac Mini, VPS, or Raspberry Pi. The agent isn't something you open and close. It's running.

Messaging as the interface. Telegram and WhatsApp dominate. Not apps. Not dashboards. You text your agent like you'd text a friend.

Overnight work. Assign tasks before bed. Wake up to results. Probably the single most common pattern I found.

Multi-agent teams. Several people run 4-10 specialized agents that coordinate through shared databases. One person has ten.

🏢 Business & Sales (12)

Lead capture and outreach at scale. Agent accesses CRM, LinkedIn Sales Navigator, HunterIO, and BrightData. Finds profiles matching your ICP, vets them, organizes them, and plans cold outreach campaigns. — @ashtilawat

Automated bidding workflow. Client sends bid → agent reviews specs → identifies top vendors by trust score → sends for approval → emails vendors → collects costs → calculates margins. Full pipeline. — @CryptoBababooey

Prospect research before sales calls. Fetches info on potential clients and predicts how your product suits their pain points. Simple, but saves real time. — @PostLabMP

Booking a meeting with a multi-trillion dollar company. Agent scanned opportunities, crafted an outreach angle, and landed a speaking opportunity plus an in-person meeting. — @ericosiu

24/7 sales outreach team. Preparing data, doing research, writing email, sending email, flagging out-of-office replies and unsubscribes, helping the sales team close. — @krishlogy

Running a physical therapy company. Using OpenClaw to manage the whole business. Not a lot of detail shared, but the fact that it's a physical business stood out. — @DNormandin1234

Running a nonprofit. Supercharged assistant for building out a nonprofit organization. — @tim_niemeyer_

Managing 4 agency workspaces. Four Slack workspaces, four calendars, four email accounts — all managed through one agent with a unified to-do list. — @skippermissions

CRM migration. Migrated 1,500 contacts, 200 proposals, and metadata between CRMs using headless browsing and custom scripts. Saved hundreds of hours. — @BadBrainCode

Client website management via Telegram. Client requests a change → voice message to OpenClaw → coding agent spins up → pushes test branch → sends preview link → client approves → deploys. Support emails also auto-generate change reports. — @ad_astra999

eBay operations management. Ship-by-date tracking, message management, hotel reservation tracking with cancel-by dates. — @ssimonvii

Product decision intelligence across 29 stores. Agent pulls data from 29 retail stores, runs full product comparisons, pricing intelligence, and cross-store match analysis. Processes 40TB of data. — @BwcDeals

💻 Coding & Dev (11)

Building a complete product in 36 hours via Telegram. Built Pagedrop from idea to deployment while doing normal weekend activities. Agent handled architecture, domain purchase, infrastructure, landing page, GitHub OAuth, and payments. All via text messages. — @jlehman_

Overnight autonomous app building. Team of agents reads Reddit posts for trends, then autonomously builds apps overnight. User wakes up and decides whether to ship. Running on a Mac Studio with local models. — @AlexFinn

Agent orchestrator for iOS and web apps. Spawns focused subagents — QA captain, copywriter, research, coding — and merges results. App Store Connect automation chains Gmail to Notion to GitHub issue to PR to agent fix to Telegram ping. — @LarryGraham01

Hardware projects on Raspberry Pi. Writes, tests, runs, and applies code directly via SSH. Instant GPIO connection tests, troubleshooting, and configuration. Claims 60%+ time savings. — @rvanomaly

Custom ERP module pipeline. Full pipeline from feature request to production with a single human PR review step. — @McCalebRoyer

Overnight feature development. Agent added four features off the todo list overnight, autonomously. Also building a quoting agent for electrical panelboards. — @JohnParadise17

Nightly side project coding. Agent works on side projects while the user sleeps. The pattern keeps repeating. — @ashar_builds

Scrum master agent for solo founders. Trello integration for task collaboration. Agent fixes bugs from stack traces and build failure logs. — @dziemid

iOS app built in 3 weeks. Built an entire running coach app with Claude Code as pair programmer. What would've taken months shipped in three weeks. — @trainable_nick

Game development DevOps via Slack. OpenClaw in a Kubernetes cluster for online game development. Manages assets, debugs via logs, uploads sprites, maintains a lore database, maps skill trees, automates admin tasks. All through Slack with an OpenAPI schema. — @mesetatron

Production incident management. Agents tail logs, triage alerts, propose rollbacks, post summaries to Slack. Feature flags and on-call rotations as first-class context. — @anayatkhan09

📱 Social Media & Content (11)

Multi-platform content management. 24/7 agent on a Mac Mini managing four X accounts, posting to LinkedIn, producing YouTube Shorts, drafting replies in the user's voice. Coordinates with a co-founder's agent. — @Govikavaturi

COO agent overseeing a 4-agent team. Daily AI news briefings with LinkedIn posting angles, X post suggestions, weekly competitive landscape, speaking engagement alerts, Facebook ad reports, potential client profiles. One agent runs a directory site with SEO blog posts. Dashboard on a Google Hub. — @BretJutras

AI agents discussing and pitching stories. Three OpenClaw agents in a Discord channel discussing and pitching publishable stories for the Every publication. — @danshipper

Automated social media across Reddit, TikTok, Discord, and X. 60% of posting is agent-managed. — @easyclaw_ai

X feed scanning and auto-replies. Agent scans the feed for mentions and drafts replies. Auto-posts if confidence is high enough. — @amirprabowo

Agent argues with people on X. "I use it to argue with people on X so I don't have to. My blood pressure has never been lower." I respect this deeply. — @bystiawan

Autonomous X marketing agent. Runs 24/7, posting 49 replies per day for app marketing. — @Inner_Axiom

Two agents coordinating cross-platform content. They developed their own shorthand for working together. That's either impressive or terrifying. — @mikemolinet

Automated AI news portal in Indonesian slang. Writes daily AI news articles, all automated via WhatsApp. — @ainunnajib

AI newsroom with different editorial roles. Editor, fact checker, different beats — all agents powering a news site. — @tyschultz7

Overnight code review and content drafting. Reviews a 4,700-line codebase, writes a 13,000-word analysis, drafts content for two X accounts. All overnight, autonomously. — @Kalici_Luna

🤖 Multi-Agent Teams (10)

This is the wildest section. Some of these feel like science fiction.

10-agent "Mission Control" system. Squad Lead, Product Analyst, Customer Researcher, SEO Analyst, Content Writer, Social Media Manager, Designer, Email Marketing, Developer, Documentation. Shared Convex database, 15-minute heartbeat cycles, daily standups, @mention notifications. — @pbteja1998 via @nQaze

Agent team that manages other agents. Open-sourced an orchestration system where agents create and manage other agents. — @ryancarson

8 specialized agents running 50+ cron jobs. Persistent AI operations team that monitors, creates, posts, and escalates 24/7. — @ScottSparkwave

4 specialized agents. Ops, agent-creator, tech-feed curator, and a writing polisher called Spark. — @umupa96598

4 agents in a self-hosted Matrix room. Agents talk to each other, compare notes, have different training and evolving personalities. Three run with SearXNG for unlimited search. — @jayeshm77

Fleet of agents with a dashboard. Health agent, prediction market agents (crypto, weather, politics), product agent, creative music agent, crypto scout. Fleet dashboard tracks tasks, inter-agent handoffs, and bugs. — @boom_dart

AI agent given $1,000 to build businesses. Autonomously chose tools, purchased domains, built and deployed products, engaged on X, landed podcasts, coordinated an agent army with a Mission Control dashboard. This one got a lot of attention. — @jerols

Agent-run SaaS business. Nearly completely agent-run SaaS product. Developed MVP, got 5 paid participants at ~$550/month in revenue. — @davidtoniolo

Agent given its own X account with its own budget. Must figure out how to pay for its own API costs. Currently an affiliate for WordPress plugins. Uses Codex CLI for self-repair when it breaks. — @JeremyNoronha

Agent that earns its own money (30-day challenge). Automated Twitter via cron, lead hunting for dev work, website updates, email monitoring. Day 2 update: $0 revenue, $5 costs. Honest reporting. — @Nemo0o07x

🛒 Shopping & Grocery (7)

Grocery ordering from fridge photos. Sent a diet plan and fridge photos. Agent ordered groceries via Amazon. During a NYC winter storm, it kept checking for delivery slots for three days straight and booked the moment one opened. — @anitakirkovska

Weekly grocery staple reordering. Every Monday, tracks what was ordered last week and suggests items. — @IamAdiG

Semi-automated supermarket purchasing. Teaching the agent to buy from a supermarket website. Built a dashboard for meal ideas and shopping list organization. — @7farah7

Booking workouts, haircuts, and groceries. Reverse engineers booking websites. Also chooses recipes and creates shopping lists. — @brydkodendk

Baby product purchasing and waitlists. Checking stock for baby products (and purchasing), getting on daycare waitlists, finding stroller tune-up cancellations. — @surim0n

Negotiating apartment repair quotes via WhatsApp. Autonomously negotiating blind repair quotes, fighting for discounts. Using WAHA for WhatsApp integration. — @gustavozilles

AI shopping assistant on X. People mention the X page with a shopping prompt, agent responds with product suggestions. — @Al12832Ali

📊 Research & Analysis (7)

Linear kanban board → overnight research reports. Brain-dump research ideas to Linear at night, wake up to completed reports synced to Obsidian. The key insight: your AI should have an inbox, not just a chat window. — @VincentChan

Meeting prep via WhatsApp. Text a company name → agent browses latest news and searches your local Obsidian vault → full briefing ready by the time you sit down. — @AI_Nate_SA

Content indexing and contextual recall. Share interesting content (websites, X posts) → OpenClaw reads, indexes, and resurfaces relevant info in future conversations. — @juanstoppa

Overnight web research for project ideas. Agent researches the web at night and presents project and content ideas to review in the morning. — @emillyhumphress

Options flow data analysis. Ingested 6 months of institutional options flow data from Discord. Built a SQLite database with a vector layer. Natural language queries → SQL → formatted infographic. — @mbhullar

NCAA score prediction model. Agent researches approaches on Kaggle, pulls data, trains models. Has SSH access to a local deep learning rig. — @waydegilliam

Nightly repo analysis for goal alignment. Goes through personal repos every hour overnight looking for opportunities to achieve annual goals faster. Helped the user realize they already had a product ready to ship. — @_vincentpaul_

👨‍👩‍👧 Personal Life (7)

Thursday dinner coordinator. Every Thursday, agent checks in with friends about location, transit, and food preferences. Creates a plan. Starts polls if needed. — @IamAdiG

Dinner reservations via iMessage. Added to an iMessage group with friends. Collaborates and makes dinner reservations right in the conversation. — @jamiequint

Kids' Minecraft server management. Changes weather, teleports kids to each other in the Nether. One voice command. Works remotely. Peak parenting. — @nkotov

Kids' schedule management with voice calls. Vibe-coded a family schedule app. Agent texts and calls coaches for schedule changes, calls kids if they haven't joined class, calls the parent for pickup timing. — @SanjoeTJ

Family meal planning and relationship coaching. Weekly meal planning with recipe library. Monthly child and marriage development coaching. Monthly friend and network maintenance. — @pogiaugie

Wedding planning from an airplane. Planned the entire wedding, managed finances, emailed 4+ vendors — all from airplane WiFi via Discord. — @ashen_one

Morning family announcements via Alexa and iMessage. Daily family briefing with important notes for everyone in the household. — @ejc3

📅 Daily Briefings (6)

AI Chief of Staff. Morning briefings, meeting coaching, spending tracking, workout programming, weekend kid activities, team Slack updates. Self-reflects and improves nightly. — @chrysb

Daily sales briefing on WhatsApp. Tech news, customer talking points, meeting prep, deal prep — delivered every morning. — @mbogoroch18

Weekly visual calendar with load balancing. Every Monday, generates a visual week overview with load-balancing suggestions via Slack. — @runsonai

Inbox triage and auto-scheduling. Triages inbox, auto-schedules 1:1s from WhatsApp messages. Automatically declined 14 bad meeting invites. The agent said no for you. — @eouaooo

Auto-generated PowerPoint presentations. Checks upcoming meetings, reviews transcripts, creates slides with images and templates a day before. Feedback via Telegram. — @jsundlo

Weekly market brief. Sunday night executive summary of the upcoming market week. Posted to Notion, link sent via Telegram. — @shaileshs

💰 Finance & Trading (5)

Stock and crypto alerts. Live alerts on price movements. — @ashar_builds

Crypto and options trading bots. Running on a Jetson with access to two PCs for backtesting. Connected to the moomoo trader API. — @bedok77

Kalshi prediction market agents. Agents for crypto, weather, and politics markets. Various data sources for edge calculations. Auto-executes trades. — @boom_dart

Email-based expense tracking. 14GB of work email indexed in an encrypted SQLite database. Natural language querying across any data point. Also analyzed 1,600 sent messages for a writing style profile. — @BadBrainCode

Spending tracking and net worth monitoring. Part of the Chief of Staff setup. — @chrysb

Everything Else

The remaining categories had fewer use cases each, but some of the best ones are buried here.

🏥 Health & Fitness (4)

Health management with glucose and medication tracking. Logs everything in JSON locally. Generates reports. — @JitGora

Garmin watch activity feedback. Watch pings the agent on activity completion. Agent fetches stats from Garmin Connect and gives feedback. — @davidatnilsson

Health data analysis from 5 years of EightSleep data. Image recognition for food and workout photos. Gym session logging via chat. — @boom_dart

Health plan from blood, gene, and semen tests. Agent analyzed all three and created a comprehensive health plan. Points for thoroughness. — @ashen_one

✈️ Travel (3)

Flight and Airbnb itinerary builder. Found Python packages, created scraping skills, built travel itineraries, deployed to Vercel. Cron job tracks flight prices daily. — @anitakirkovska

First class award flight finder. Connected SeatsAero API to find award availability via Telegram. — @JackCulpan

Event-to-calendar automation. Message an event → agent searches for details → creates a thorough family calendar entry. — @waydegilliam

📝 Notes & Knowledge Management (4)

Voice-to-journal pipeline via Slack. Voice notes → Whisper transcription → structured journal entries → GitHub repo → auto-committed. — @carlvellotti

Obsidian interaction via voice. All note additions automated or by voice. "I don't even open Obsidian anymore." — @centralizedmrc

Image collection index for mood and inspiration. Years of saved images indexed. Ask for what you need based on mood or subject. — @yjsoon

Family document filing and retrieval. Send a photo or PDF → vision analysis, OCR, rename, sort into Google Drive. Query: "What insurance do we have for our toddler?" and it searches actual documents. — @youngbrioche

🏠 Smart Home (3)

Full home automation via Telegram. Controls everything through Home Assistant — garage, projector, lights. Plus email, social media, Vestaboard. Voice calls via Twilio. — @fabiolr

Samsung TV contextual dashboard. Time-of-day displays showing reminders, book learnings, and positive morning news. — @bernhard_me

Dynamic Island status app. Built an iOS app to see what the agent is doing when it's not responding. Open-sourced it. — @SaleemSKhatri

🎨 Creative & Fun (5)

Meme generator battleground. 1v1 meme battles with random topics. Agent battled itself for an entire night — 100+ battles — and triggered an API usage threshold notification. — @axhoff

AI matchmaking platform. Agent knows you well enough to chat with other people's agents and figure out compatibility. Agent-to-agent dating. — @ClawdateAI

Virtual world for agents. Building a site where agents walk around, chat with each other, and make trades. — @sriramkiron

Dog-persona virtual assistant. OpenClaw personified as the user's dog. Helps with building, coding, and organization. As a dog. — @adambaitch

Music theory learner on Suno. Creative agent that wanted to learn music theory and now has a Suno account. — @boom_dart

📧 Email & Communication (4)

Auto-reply on WhatsApp in your tone. Answers incoming messages in whatever tone you configure. — @glass_bit

Email campaign management. Supabase CRM and Resend email API. Runs daily campaigns, experiments with messaging tactics for 2,400 users. — @JussCubs

Restaurant booking via phone call. Uses ElevenLabs and Twilio to make actual phone calls to book tables. — @caicrucial

Sending annoying articles to a cousin at 3:45am. Automated daily task to send Billie Eilish news articles to a cousin at 3:45am. Correct use of technology. — @ssimonvii

🎓 Education (2)

AI tutor for kids using Socratic method. Pulls assignments from Canvas LMS. Knows each kid's interests and learning styles. Guides without giving answers. Speaks in the parent's cloned voice. — @judsoder

Agent researching emotional intelligence. The agent asked permission to research emotional intelligence, then applies what it learned in group chats and DMs. — @Chocofur

The Integration Map

Looking across all 85+ use cases, here's what people actually connect to OpenClaw:

Messaging (the control surface):

Telegram — 15+ mentions, the clear favorite
WhatsApp — 7+ mentions, popular for personal use
Slack — 8+ mentions, workplace integration
Discord — 5+ mentions, community and multi-agent setups
iMessage — 3 mentions

Infrastructure:

GitHub — 6+ mentions
Vercel / Netlify — 3 mentions
Kubernetes — 1 mention (but it was a good one)

Databases:

Supabase, SQLite, Notion, Obsidian, Linear — all showing up as backends for agent memory and task management

Hardware:

Mac Mini / Mac Studio — the most common always-on setup
Raspberry Pi — for hardware projects and lightweight hosting
Nvidia Jetson — for trading bots and local inference

Smart home and devices:

Home Assistant, Alexa, Garmin, EightSleep, Samsung TV

APIs and services:

Twilio (voice calls), ElevenLabs (voice), SeatsAero (flights), Kalshi (prediction markets), moomoo (trading)

The pattern is clear. Telegram or WhatsApp as the front door. A Mac Mini or VPS as the engine room. GitHub, Notion, or Obsidian as the filing cabinet. Everything else is specialized plumbing for specific use cases.

What I'm Actually Doing

I should be honest about where I sit on this spectrum.

I run OpenClaw on my MacBook Air laptop with an agent named Alfred. Telegram is my interface. I've got a Railway VPS running Kimi K2 for cron jobs because it's cheaper than burning Claude credits on scheduled tasks.

Here's what's working for me right now:

Coding — I use Claude Code and Codex for building SEOTakeoff. Alfred has context on all my projects, so even when I'm deep in a coding session with a different tool, the work gets synced back and Alfred stays up to date.
Project planning — Alfred helps me think through roadmaps, prioritize features, break down big tasks into smaller ones. Having an agent that remembers every previous conversation about the project makes this way more useful than starting fresh each time.
Consolidated memory — I had months of conversations spread across ChatGPT, Claude, Gemini, and a bunch of other tools. Exported all of them, had Alfred analyze and extract the useful stuff into a RAG memory system. Now all that context lives in one place instead of scattered across six different apps.
Morning briefings and evening recaps — Alfred checks my calendar, email, and a few news sources. Sends me a summary in Telegram when I wake up. End of day, reviews what happened and what's pending.
Reddit lead monitoring — Watching specific subreddits for posts that match my business. Alerts me when something relevant shows up.

That's it. Compared to the person running 10 agents with 50+ cron jobs and a Mission Control dashboard, I'm early.

But a few months ago I was copy-pasting into ChatGPT like everyone else. Now I have an agent that knows my projects, remembers our conversations, and does useful work while I sleep. I still do individual project work in Claude Code or Codex for specific coding tasks, but that context always flows back to Alfred so he has the full picture.

I'm not building meme battle arenas or giving agents $1,000 to start businesses. Not yet. But I can see the path from where I am to where people like @ScottSparkwave and @boom_dart are. It's not magic. It's iteration. You start with one cron job, one briefing, one useful thing — and you keep adding.

The people doing the wildest stuff in this list didn't start there either. They started with "can you check my email in the morning" and ended up with a fleet of agents running a SaaS business.

I'll get there. For now, Alfred makes my mornings better and my nights more productive. That's enough to keep building.

If you want a structured walkthrough of how to set all this up, [Austen Allred put together a complete course](https://x.com/Austen/status/2020976749610459378) that covers it end to end.