muse-wonder

Grok Build, Muse Code, Claude, Codex & more in one place; GUI; AFK Pilot

One brief in, a whole course out: Meta Muse agents write the plan, a lecture per topic, pictures and exercises; code checks every quote and grades every answer.

Three AI roles (Proposer, Critic, Evidence Evaluator) argue an agent's proposed change on Meta's Muse models. The runtime verifies every record's proof, every event is hash-chained, and only a person decides. Meta Global AI Developer Hackathon 2026.

Translucent is the intelligence layer between scientific observations and physical experiments. It reads what the lab measured, learns how the system behaves and designs the next experiment, accelerating discovery in frontier materials and energy research.

AI Social Media Commercial Studio (BYOK) - 10s, 20s, 30s, 60s commercials for social media platforms. No subscription. Built during the Meta Global AI Developer Hackathon 2026

🛰️ OceanEye Sentinel 2.0 — Verification Kit & Meta AI Benchmark Harness: Sentinel-1 C-Band SAR, Meta SAM 3.1 Gate A Segmentation, Muse Wonder 1.0 vs Muse Spark 1.3 Arena, and RFC 8032 Ed25519 Chain of Custody.

Photo-based crop analysis for Kenyan farmers, built on Meta SAM and Muse, in English and Kiswahili

Meta AI Developer Hackathon

Meta Global AI Developer Hackathon (Indie Track - Push the Frontier): Zero-CSO wastewater MPC uniting 6 Meta models (SAM 3.1, Muse Spark 1.3, Muse Wonder 1.0, Muse Voice 1.0, Muse Image 1.0, Muse Glimmer 30B) with EPA SWMM5 DYNWAVE & SHA-256 verification.

Stretto: a local time, focus and efficiency tracker that finds where your day was lost and turns it into ranked changes with Meta Muse models. Humanflow is its voice and agent backend. Meta Global AI Developer Hackathon, Build for Your World.

Point your phone at a broken machine: Helpa reads your exact model and shows the fix on a 3D twin, then checks your work. Meta Muse Spark, Muse Wonder, SAM 3.1, Muse Image. Built for the Meta Global AI Developer Hackathon.

Evidence-backed object histories from short fixed-camera video. Follow one object, answer questions with claims linked to the original frames, and say so when the footage doesn't show what happened. Built on Meta SAM 3.1, Muse Spark and Muse Wonder.

Evidence-first verifier for AI-generated pull requests. Muse Spark writes tests that must pass on the original code; each runs 3x on base and patch in a no-network Docker sandbox, Muse Wonder checks Spark's work, and plain code decides PASS / REVIEW / BLOCK. Built for the Meta Global AI Developer Hackathon 2026.

QorTime AI: Meta Muse Spark, Muse Wonder and SAM 3.1 explain, red-team and visualize a deterministic video-game integrity referee (anti-cheat + verified skill, signed match receipts). The AI never decides a verdict. Meta Global AI Developer Hackathon, Indie Developer track.

RoboWitness — independent verification for AI robotics fixes, powered by Meta AI. Meta AI proposes, deterministic PyBullet sim executes, oracle decides. Sealed SHA-256 evidence.

Attma: a research agent on Meta's Muse models that turns a biology question into a reproducible, checked workflow.

Factory disruption recovery with source-linked evidence, Meta Muse Spark Open on Baseten, deterministic capacity planning and human-authorized decisions. Fireworks standby.

Solidity security verification powered by Meta AI: Slither detects, Muse Spark Open explains and ranks findings, Slither re-checks proposed fixes, and optional Muse Wonder retries a rejected patch up to three times. Built for Web3 developers and non-technical founders.

Supervisory layer for AI coding agents: parks approval requests, keeps independent work moving, detects repeated failures, gets a Meta AI diagnosis, and verifies recovery with real test results. Go + SQLite.

JYNTRA AI - Your AI. Your Edge. An AI-powered financial intelligence platform combining Meta AI models, real-time market data, and a six-agent analytical system to deliver structured market insights, risk assessment, and trading decision support 📈.