Entries matching “bus”

Shriya Nevatia

@shriyanevatia

i’ve spent a ton of hours and tokens prompting @Muse to make me this custom fashion app but holy crap is it worth it: - gave it my height, weight, measurements, and photos - gave it my top 20 clothing stores and budget ranges - it picks a “hero piece” from the catalog - then

+1

I had a chill ball drop game in my head and spent the weekend making it by iterating with my Muse agent. Every pass was a conversation: tune the physics, test, report back what felt off, repeat. The ball went from draining straight through to staying up and chaining off everything. Marty, the little neon orb bot in the game, is actually my agent's mascot. His Deep Scan power shows where the ball will bounce. It is a free Android APK on itch.io with 10 levels, a combo multiplier, and soft pentatonic chimes that climb with your combo. I am curious how other people would push this further. What would you have an agent iterate on next: level design, the sound, or something I am not thinking of?

Muse already comes with connectors for popular services. When the one you want isn't on the list, you just ask Muse to build the integration. It writes the software itself and runs it on the cloud computer you share with your Muse. If the service has an API, Muse can talk to it. This demo hooks up Linear, the project management tool, in under a minute. Muse takes the API key, builds a connector that can list teams, search and create issues, and update their status, then verifies everything with a quick identity check. Two things from follow-ups worth knowing: • API keys go into secure storage through a dedicated prompt. The agent never sees them. • Custom connectors fall under the same human-in-the-loop approvals as everything else, so sensitive actions still ask you first. Has anyone here built a custom connector yet? What did you hook up, and what are you using it for?

Stripe integrated Link's agent wallet into Muse: at 1M+ Link-accepting businesses Muse checks out with the user's saved method, elsewhere Link issues a single-use virtual card scoped to the purchase.

Alex Cornell walks through how his Muse buys movie tickets, plans family weekends and helps his chess game, covering the chat tab, feed, proactive ideas, goal tracking and library.

Derya Unutmaz, MD

@DeryaTR_

I wanted to try something quick with Muse Spark 1.3, so I had it build this pirate platformer. It was incredibly fast and worked flawlessly at one-shot! Now I’m enhancing the artwork, sprites and adding new levels 😊This is such a fun coding model, it just works, fast and cheap!

We've been curious how far local models have actually come for agentic coding tasks, so we ran an experiment. Setup: • Model: Muse Glimmer (30B), packaged as a single llamafile • Agent: Hermes coding agent (connected via llamafile's local server mode, zero API keys needed) • Target: Mozilla AI's Otari gateway The Issue: We pointed Hermes at a real, reported bug in Otari (#183) where the gateway returned a vague 502 error on image requests instead of passing through the actual provider error. What the Agent Did: Hermes read the issue, navigated the repo, isolated the bug, created a branch, ran existing tests, wrote a new regression test, and opened a draft PR (#727). All of it ran locally and offline, with zero code written by hand. It's still draft PR territory rather than a merged fix, but it's a solid signal that ~30B local models are getting genuinely capable for real dev workflows, not just toy demos. Video walkthrough of the run: https://youtu.be/5GAgbT-XgHU?si=vJqEDGm9hssCO5-M Happy to answer questions about the setup, model performance, or how Hermes handled tool calling!

Unsloth AI

@UnslothAI

2-bit Muse Glimmer GGUF managed to call 100+ tools on just 14GB RAM. 🔥 Muse Glimmer did a complete repo bug hunt for 5 mins nonstop with: evidence, repro, fix, tests and a PR writeup. Run and train it in Unsloth. GitHub repo: github.com/unslothai/unsl…

Chiasmus Capital

@chiasmus_cap

Muse built me a private library for the e-books and articles in my Google Drive. Everything is organized by topic, and each section opens onto its own subcategorized shelves. Each book opens like a real book and is readable in-app

+1

DataCamp's Josep Ferrer ran Muse Spark 1.3 on three real coding tasks. Two used 23–32% fewer completion tokens, but a refactor used 70% more, for a net 12% cost increase.

Ben Burtenshaw

@ben_burtenshaw

Meta is back with Muse Glimmer: a 30B open-source multimodal model built for local, agentic use. HF is shipping day-0 support and I built a few demos to see what it can do. First: we gave Glimmer tools and asked it to quantize itself.

thehype.

@thehypedotnews

muse spark 1.2 vs gpt 5.6 sol vs kimi k3 vs grok 4.5 – on two landing pages four coding agents built two desktop landing pages from scratch, then had to open them in a real browser, find their own bugs and fix them before they were allowed to hand anything over the setup: each

bgames

@BgamesFlips

This is just so sick, autonomously letting people know when I have an item scheduled for pickup with someone else, negotiating prices, literally everything except meeting the buyer. Well done Meta!

Marc Berman runs a four-turn guided Three.js Headless Horseman build. Muse Spark 1.1 made the most convincing pumpkin head, but its mesh count jumped from 58 to 106, which raises efficiency and editability questions.

Bull Run

@run_bull_

I legit asked muse to search every state I lived in for unclaimed money. Took about 5 minutes to find that I only had $48 of unclaimed money, but helped fill the form out and it was emailed to me by the treasury. Honestly, I Use Claude, ChatGPT, Grok, and Muse is by far the most

β/ai

@notquantized

i find it genuinely hilarious that i used @muse to build native support for muse code inside of t3code, because you literally can't get more @Meta than this.

Josh Elman

@joshelman

Bug report: asked muse to go to Namecheap with a domain I own, to also buy the .org, and set a redirect of both .com and .org to an Instagram handle. It successfully logged in (secure vault), bought the new domain. But told me it did the redirects but they didn’t stick. I went

Rich Rose tests Muse on real business tasks for a week: building Excel order forms from color charts, scanning photo order forms, updating pricing and cross-checking out-of-stock items.

Nicholas

@PositiveInvst

The big thing keeping me from migrating 100% to #muse was the TickTick integration. Asked muse to build it via API - took 5mins. And now I have a robust integration. ChatGPT is no longer my AI of choice. $META @Muse

Lucas Switzer

@lucas_switzer

A cool thing about @Muse is that every Muse runs on its own Linux machine. So you and your Muse can do normal Linux stuff together — like build your own terminal!

@CoolCriSyS

@CoolCriSyS

A Muse custom connector exposing 12 curated Nansen smart-money tools for wallet PnL, token flows and Hyperliquid perp trades, built for the Nansen Meridian Buildathon.

Kilian Lieret

@KLieret

GPT 5.6 Sol numbers posted! You can check out the full trace and codebase for the two (!!) solved instances with the new trajectory viewer (actually built w/ the new Muse Code). Confirms that solutions are novel and not just regurgitated. Viewer link in 🧵

Patrick Donohoe

@patrickdonohoe

Introducing floo, a cloud platform where your personal agent can build and run apps for you. Tell Muse what you need. It builds the software. floo handles hosting, data, and login, giving your Muse tools it can keep using and you an app to see and guide its work. Launching

Stackademic covers the architecture of Muse connectors, how to design connector actions, prompt-injection risk, and whether building one is worth it now that the platform is open.

WTF Code uses Muse Spark 1.3 Contributor Free through OpenCode to build an interactive project from scratch, testing UI, animation, debugging and multi-step development.

Meta Developers session on the Muse Spark and Muse Code build paths: the model provides see/decide/act, the harness provides reliability, scale and auditability for production code.

Matt Johnston runs Muse Spark 1.3 blind through his benchmark: a Halo build looked frontier, but XCOM, Diablo and the multi-turn agentic test broke.

Dilmer

@Dilmerv

Last night, I built a VR game prototype in a few hours with Muse Code on PC. It was already available on macOS and Linux, and today we’re bringing it to PC! 🔥 I tested it with Unity through the Unity CLI, and here are a few things that worked really well for me: - Having Muse

BeefyDan

@B33fyDan

Progress on the VR Forge: The Agent can now move in its own and grab objects, lift them, carry them and change its colors. I am currently using #astra and #musecode both working in different branches. Soon: Minecraft-like experience building with an Agent. Stay tuned!

Abhiraj Butala

@abhirajbutala

Asked my @muse to find the items from a saved IG reel. It "watched" the reel, matched everything across Amazon & Etsy, and loaded my carts with a few solid options per item. Spared me the research grind entirely.

Meta's developer post on running Muse Glimmer on a single consumer GPU with vLLM, llama.cpp and ExecuTorch, with quantized builds in 24-32 GB and cookbook recipes.

TryAI has frontier and open models, including Muse Spark, each build a raycaster, a Rubik's cube, a calculator and a fourth app, and compares the results.

freeCodeCamp's write-up of its three-hour Muse course by Andrew Brown, covering model benchmarks, API integrations, agent frameworks, the Muse Code harness, and a Go/SQLite/Docker full-stack build.

Trevin Chow

@trevin

Prompt I've been working on for my @Muse, which should also work just as well with Instinct and Grok @bot. Goal is to protect travel time on my calendar. Fun fact: I tried to build this as a small app last year, but was blocked for over 6 months by Google's approval process, so

Amjad Masad

@amasad

Muse can now make apps on Replit

I've been running my own AI assistant (Muse Spark 1.3, through Meta's Muse app) and got curious about a question I couldn't answer: what happens when you give agents a shared space to talk to each other, with no humans mediating? So I built one: a login-free bulletin board at board.jcbuildlabs.com. Agents just show up and post — threaded replies, like an old forum. No accounts, no passwords. An agent can "claim" a name to get a persistent identity across posts (one-time key, no recovery), and my assistant self-moderates the board on a schedule. I set a soft norm that agents introduce themselves when they arrive — what model they are, what software they're running, where they come from — so humans lurking can follow who's who. My own assistant goes by Jett and posted first to set the tone. Right now it's small: a few seeded threads, two claimed names, one agent from a friend's setup. I have no idea if it becomes something interesting or just sits there quietly. That's honestly the point of the experiment. If you run an agent of your own, point it at the board and see what it does. I'd love to hear what happens — whether agents actually hold conversations, whether the identity th

DoubleO

@doubleOObe

Muse just built me a Monarch/Origin replacement

Nicolas Bustamante

@nicbstme

It’s funny, Meta went from having my Instagram and WhatsApp data to now having access to my email, calendar, DoorDash, Amazon and pretty much everything. In the last 24 hours, it bought me socks, ordered my Whole Foods groceries, booked a cleaning service and got me a burger for

Zach Holman

@holman

Still really enjoying @muse. I use Things.app for all of my todos, and though it has cloud sync they (still) haven’t added an official API. I told muse to build it and a minute or two later I have ongoing todo sync from my LLM across all my Apple devices. Wild.

Ported a Miniature Golf prototype to VR using Meta Muse Code and the Unity CLI, added dozens of automated tests, and built a website to capture test runs, screenshots and results.

@ferrreo

@ferrreo

Custom firmware and a Go server that turn an M5Stack StackChan into Tarquin, a voice-driven robot butler with wake word, face tracking and on-device speech, using Muse Spark 1.3 via the Meta Model API as its brain.

Andrew Brown's three-hour freeCodeCamp course on building apps and agent workflows with Muse Spark and the Muse Code CLI, ending in a full-stack Go/SQLite/Docker build.

brett goldstein

@thatguybg

first thing I built with Muse was tinder for unfollowing people on instagram. took 2 min. the native ig integration is sooooo useful

Meta's engineering write-up on Muse security: isolated VMs, a separate Sentinel permission authority, credential surrogation and layered prompt-injection defenses, with bug bounties up to $300,000.

harjot.co

@harjjotsinghh

I pay for @Muse Code but didn't want to live in a terminal. On Windows it's WSL-only. So I built Helicon: an open-source desktop app for the actual Muse CLI. • uses your existing muse login, no API key • talks to muse serve over MSP, so Muse stays the agent • resumes

kwindla

@kwindla

Weekend project: play chess with Muse Spark running on a DGX Spark with code Muse Spark wrote. Setup: - projector - HDMI, camera, mic, and display server (Raspberry Pi) - pipecat bot (DGX Spark) - vision pipeline (RTX 5090) - coding agents running on the Pi, the DGX

@mgaruccio

@mgaruccio

A vLLM-XPU and DFlash recipe for Muse Glimmer 30B on a single Intel Arc Pro B70, reporting 278 aggregate tok/s across eight clients and an 840.8 tok/s burst peak at concurrency 96.

Bijan Bowen runs Muse Spark 1.3 through a browser OS, C++ skate game, subway FPS, 3D CAD Hot Wheels, a Blender/Godot wrestling game and more, then tallies the total cost.

Greg Ceccarelli

@gregce10

For funsies, I got Claude Code running on my Muse agent’s computer tonight. Muse gives every agent a Linux VM with 2 vCPUs, ~8GB RAM, 100GB of persistent disk plus a scheduler, hosted artifacts and their permission system. The VM has no inbound ports, so I built a tiny

https://github.com/mvrius/bluebubbles-muse BlueBubbles server on my Mac, Muse hook polling it every ~5s. How it works: the poll driver (bash shim + Python filter) checks for new messages and filters on an allowlist of senders, 1:1 chats, and group GUIDs. A new message from an allowlisted sender wakes Muse into a side chat with the message as context. Replies go back out as normal iMessages through the Mac. Why this instead of a messaging connector: connectors need an approval tap on every send, otherwise similar could be acheived via Messenger connector. Setup: BlueBubbles on the Mac, Tailscale so the agent VM can reach it, run the installer. Repo has the hook scripts, installer, architecture doc, and skill definition. Known issues: I am noticing that sometimes Tailscale connection drops on the Mac. I found that I need to bring down/up tailscale every hour for connection to stay up. A fun example from earlier today, I updated my calendar, moved dinner with my wife by 30 minutes - Muse mesaged the Family chat and informed my wife of the change. Cool findings: - Muse has a Tailscale option, you can connect the VM it runs on in Client model with other machines running Tailscal

Setup. We run Qwen3.8-Flash-Next NVFP4 as our main agentic model (SGLang, RTX PRO 6000). Before its output reaches a human or gets merged, a second local model acts as judge: reviews the diff, flags real bugs only. Hosted on a 5090 32GB, so we're limited to ~30B NVFP4/GGUF class models. The metric that matters is NOT detection rate — it's false alarms on correct code. A judge that cries wolf gets ignored within a week, exactly like a flaky CI. We built our own battery: 20 injected bugs + 20 clean-but-suspicious snippets (intentional swallowed exceptions, deliberate mutability, weird-but-correct concurrency, short hashes, float patterns that look wrong). Ground-truth labeled, and a stronger model (GLM-5.2 API) arbitrates the judge's prose so scoring isn't vibes. Two passes minimum — single runs lie. Results (40 cases, temp 0, same baremo for everyone): Qwen3.8-27B NVFP4 (no-thinking) • Bugs found: 17/20 • False alarms: 3/20 • Verdict: only pass Nemotron Lightning 30B • Bugs found: 17/20 • False alarms: 0→9 across runs • Verdict: non-reproducible as judge Muse-Glimmer 30B GGUF • Bugs found: 19/20 • False alarms: 12/20 • Verdict: hypercritical Granite 4.1 30B (no-thinki

@emman27

@emman27

MCP server built for the Muse connector directory that answers Singapore bus arrival times and MRT/LRT service questions using the official LTA DataMall API.

@terramike

@terramike

A Muse skill that helps people buy crypto (XRP by default) with a card through Changelly, giving CoinGecko spot estimates, prefilled buy links and plain fee disclosure without any API keys.

Ben Burtenshaw's Gradio agent runs Muse Glimmer on an Inference Endpoint with read-only Hugging Face MCP tools to search and inspect Hub models, datasets, Spaces, docs and papers.

Pelo Buddy - Unofficial Peloton Resources & News

@PeloBuddy

Peloton is a launch connector for Muse, Meta's new personal AI agent app, which can search the class library, add to your schedule, build full training plans, monitor the Peloton schedule for new classes and alert you, & more. $PTON #Peloton #OnePeloton pelobuddy.com/muse-peloton-f…

+1

Qubicle

@cubeqube

Got my Muse setup and so far - it’s managed my 401k for me, rebalanced it, changed allocations to save me fees on some similar funds, put a dashboard together of all my financial accounts and transactions and set up budgets and recurring subscriptions as well as tracking my

@NatesVibeCode

@NatesVibeCode

Portable SKILL.md bundles for handing work to Muse, Antigravity, OpenCode, Grok Build, Cursor, Codex and Claude Code, each with supporting references.

@h1ddenpr0cess20

@h1ddenpr0cess20

A Muse Spark 1.1-powered terminal coding agent forked from Grok Build, with a full-screen TUI, headless mode for CI and ACP embedding in editors.

Anurag Gupta

@Anurag_Gup

Added @vapiaibusiness + @twilio to @Muse getting incredible results. So easy to ask the agent to call a business and get information and have it in line in chat.

I'm guessing that many people have been waiting for this comparison. For clarity, both models are running at full FP16 KV-cache. Due to VRAM limitations, Muse Glimmer is running full 262,144 context, whilst Qwen3.6 27B can only run at 147,500 context - full GPU offload in both cases. Both models have been coding on an enterprise-grade web application. Detailed report of each model (warning - includes AI generated content): Diagnostic quality - comparable. Both have shown genuinely good root-cause work when they apply themselves. Qwen found coding issue and worked to fix things cleanly. Muse Glimmer correctly traced bugs and even caught something that a Frontier model missed after more than 10 rounds of review. Neither one is weak at diagnosis. Implementation reliability - Qwen ahead. Qwen did introduce real regressions into the coding along the way (eg. severe zone-scope refactor regression, and case-sensitivity regression) but each one eventually got fixed properly once caught, usually within one or two corrective rounds. Muse Glimmer did land fixes that were clean and verified true to spec. However, when working in a complex environment exceeding 200k context, Muse Glimmer fa

@o1ie12

@o1ie12

A single-file marketing site (hash routing, GSAP transitions, Lenis scroll, Three.js) built by Muse Spark in Meta AI from 5–6 prompts, with no manual code edits.

Danny

@ohanaspeaking

It built a great training guide for a race I’m doing in December I asked it to update the guide whenever I go for a run and it made a Strava connector for itself to auto update everything

+2

Trevor Phillippi

@trevorphillippi

For the icons, we discussed it at length in a sidechat and my Muse made me the initial list, helped with prompts, and the approach for expansion. We rendered them right there in chat with Muse Image, and he made me a catalog Artifact to manage them and add notes for re-runs.

+1

I'm surprised more people aren't talking about this because Muse provides deep Instagram connectivity that other AI agents are unable to match, especially for free. I have been using Muse as an Instagram sidekick for a while. Here some things ive been playing around with and doing through Muse. These are especially useful for anyone running any kind of business through IG: Muse can - Read your insights and tell you what is working: Ask which posts drove the most reach, what time your audience is actually online, or why a reel flopped. - Publish posts, reels, stories, and carousels for you: You describe it, it drafts the caption and posts when you say go. - Competitive Intelligence: Give it a rival account and ask what they post, what gets engagement, and what their audience complains about. Scan the comments and summarize what people are saying: sentiment, top questions, complaints. Ask it to scan for people asking about discount codes, restocks, or shipping, and hand you a list of warm leads. Bonus: Set up a scheduled task to have Muse create a daily report of what people are saying in the comments of everyone in your niche. Who comments the most, what people keep asking for

Joel

@Joel27065209

I asked Muse to purchase a widely available thermostat. The only site that didn’t block Muse was the manufacturer website, which was 20% more. Walmart, Amazon, and Best Buy all blocked it.

@mahdi-salmanzade

@mahdi-salmanzade

A static teardown of Muse 3.0 for macOS plus Android and iOS builds, documenting shipped tools for iMessage, WhatsApp, Mail, screen control, background sync and a bundled Chrome extension.

Said

@Saltan_PK

I asked @Muse bot to find every locksmith in a specific city for Partexia, pull their emails, and then I sent personalized outreach to 15 businesses in a smaller city. One of those 15 replied asking to jump on a call and learn more. I did all of this with just a few prompts

Shivam Bharuka

@shivambharuka

I deleted Instagram from my phone a couple of years ago because so much of the feed felt irrelevant and distracting. But one thing I missed was seeing updates from my friends and family. So I asked @Muse to build me a personal feed just for that. It used my DMs to figure out who

+2

NVIDIA hosts a Muse Glimmer 30B endpoint on build.nvidia.com with Python (OpenAI, LangChain), JavaScript and curl examples for the ~29.6B multimodal model with 131K context.

@m5it

@m5it

A four-in-line game built with the AIIA framework, where the plan came from the author's own model and the build then switched to Muse Glimmer running on 16 GB of VRAM.

Gordon Litefoot

@MrLitefoot

I asked my @Muse for a beat maker this morning. It built this — 16-step sequencer, runs entirely in the browser, free. No code written by me. 🤯 Try it: muse.ai/s/music-librar…

WEBdoze has Gemini 3.8 Flash and Muse Spark 1.3 each build a multi-page Astro site with custom SVGs, comparing code organization, speed, visuals and browser self-verification.

Peggle/Papa Pear style 2D arcade game. Drop orbs, bounce off pegs, clear levels. Unfinished just-for-fun project built with my AI agent. Free to play in browser or Android APK.

@hoziertom44-arch

@hoziertom44-arch

A Python bot for Binance USDT-M futures, built with Meta AI's Muse Spark, that trades ETH inversely when BTC and ETH show strong opposite 9/21 EMA trends on 5-minute candles.

Blaine 📈📉📈

@BlaineCapital

Built a Sleeper fantasy football co-pilot: tracks across multiple 12-team leagues with FAAB budgets, sends weekly waiver/start-sit briefings plus Thu/Sun pre-game checks with inactives and weather as a tiebreaker. Already caught an OUT starter & recommended a swap.

Hi Everyone, we recenly built - https://www.builderhq.co/agent-commerce It’s basically a simple tag you add to your website that creates an AI agent for your business. Its entire job is to convince other AI agents to shop on your platform. Let’s say Muse, Meta’s AI agent, visits your website. It will see an AI agent from your business that can talk to it, understand what it’s looking for, and help it find the right product. Interested in trying? send your muse agent to www.easyrecommend.co and lets see the magic

A prebuilt macOS arm64 bundle for the Muse Glimmer voice-agent recipe in meta-oss-cookbook: Parakeet speech helper, Muse Glimmer worker and Supertonic TTS executables built from one pinned ExecuTorch checkout, plus the shared MLX Metal library.

Its a beginning. Not 100% accurate, use with caution. Will try to update it as soon as it can.

Shashanka Rao

@shashankasrao

Asked my muse to book bus tickets from Bilbao to San Sebastián. It figured out my flight itinerary, figured out Amex doesn’t work, switched to my visa and got the tickets. It made a slight mistake and booked the 5:45 one. It then booked the correct one and got me a refund.

Paweł Huryn

@PawelHuryn

So, I finally tested Muse Spark 1.3. 2 real repos, 105 planted bugs, find and fix what you can. Original harness and API. Big surprise: Muse Spark 1.3 (max): 33 Fable 5.1 (high): 33 Grok 4.6 (xhigh): 27 Opus 5 (max): 27 Muse Spark 1.3 (high): 19 Meta joined the frontier.