# Hack FLUX: Beyond One: Project Gallery

- **Event:** [Hack FLUX: Beyond One](https://cerebralvalley.ai/e/bfl-hackathon)
- **When:** Nov 22 at 9:00 AM – Nov 23 at 5:00 PM (PST)
- **Where:** Shack15, San Francisco, CA
- **Hosts:** [Black Forest Labs](https://cerebralvalley.ai/u/bfl), [Cerebral Valley](https://cerebralvalley.ai/u/cv), [Shack15](https://cerebralvalley.ai/u/shack15)
- **Projects:** 29 (6 placed)
- **Page:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery

## Projects

### 1. BuildViz

BuildViz generates photorealistic architectural renderings with itemized construction cost estimates in 60 seconds, powered by FLUX.2.

Contractors, architects, and developers need feasibility studies to win bids, but traditional renderings cost $5,000 and take 2 weeks. BuildViz uses FLUX.2's photorealistic capabilities to create renderings that don't look AI-generated, combined with Claude AI for cost estimation and NVIDIA Nemotron for permit analysis and multi-interpretation design workflow.

Problem Statement Two: Photorealistic Detail

- **Team:** [Yi Zu](https://cerebralvalley.ai/u/yizucodes)
- **GitHub:** https://github.com/yizucodes/BuildViz
- **Demo video:** https://youtu.be/F0rx_DTFUo8?si=pcK2eNWiCNOx5648
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=1

### 2. Exitflow

Architects currently test evacuation flows using abstract dots on 2D maps. We believe safety planning deserves better—it should be visualized, photorealistic, and actionable.

We are solving Problem Statement One: Cross-Image Consistency. Our engine uses Fal AI to generate completely different crowd behaviors while keeping the original building geometry pixel-perfect across every scenario.

We upload one facility photo. Scenario A generates a calm exit flow. Scenario B simulates a high-stress emergency evacuation. The lighting, textures, and architecture remain identical—only the human behavior changes.

This allows safety planners to stress-test infrastructure visually before construction begins. With ExitFlow, we aren't just generating images—we're designing safer exits and potentially saving lives.

- **Team:** [Lily Su](https://cerebralvalley.ai/u/lilyxsu)
- **GitHub:** https://github.com/LilySu/ExitFlow
- **Demo video:** https://youtu.be/1D55OPRQC2M
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=11

### 3. Flipbook

A flipbook using generated images. Upload your own drawing and a prompt and the drawings generated will use your style consistently.

- **Team:** [Anna Saltveit](https://cerebralvalley.ai/u/BananaDuck)
- **GitHub:** https://github.com/asaltveit/Flipbook
- **Demo video:** https://drive.google.com/file/d/1p_v-1c9PbVzYC6lSmS57wo0fypRrbId9/view?usp=sharing
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=15

### 4. InteriorFlow

InteriorFlow solves the universal anxiety of preparing your space when hosting guests. Whether it's a holiday gathering or casual get-together, most people stress over how their home looks and what decorations to buy.
Our solution:

AI-Powered Room Redesign: Upload a photo and chat with AI to generate photorealistic redesigns for any occasion. Visualize changes before making purchases.

Smart Shopping: Automatically scrapes and curates decoration products from Amazon for one-click purchasing.

Spatial Intelligence: Computer vision extracts and measures furniture, windows, and room dimensions to ensure suggestions actually fit your space.

AI Design Advisor: Identifies layout issues, suggests improvements, provides measurements, and recommends specific products—expert guidance without the expert price tag.

- **Team:** [Kaarel Kaarelson](https://cerebralvalley.ai/u/kaarelkaarelson), [Sajeev Magesh](https://cerebralvalley.ai/u/sajeevm), [Andrew B McMorrow](https://cerebralvalley.ai/u/AndrewMcMorrow)
- **GitHub:** https://github.com/AndrewMcMorrowGlitch/BFLCerebral
- **Demo video:** https://drive.google.com/file/d/1tksDH4gSsCbZ9TkntmsCRefsZafXWBjd/view
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=17

### 5. HeroBoard

HeroBoard turns generative models into a real storytelling medium.

HeroBoard is a multiverse storytelling engine that lets anyone create branching comic narrative videos with consistent heroes, worlds, and props. You start by uploading a hero (e.g., Batman) and a few key assets (e.g., Gotham skyline, Hogwarts Great Hall), then build out the story in a visual workflow where each node becomes a new comic panel. HeroBoard auto-suggests in-character next moves (i.e. it creates the story with you), keeps the look consistent with reference-guided image generation, and reuses props/locations from a Side Asset Library for fast worldbuilding. The magic moment is you can make it into videos.

- **Placement:** Finalist
- **Team:** [Vedika Jain](https://cerebralvalley.ai/u/vedika), [Janani Prasad](https://cerebralvalley.ai/u/jananip), [Purav Patel](https://cerebralvalley.ai/u/puravpatel)
- **GitHub:** https://github.com/PatelPurav05/HeroBoard#
- **Demo video:** https://www.loom.com/share/55d9e4aeb1ce4a7c82ed9d74100223d1
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=23

### 6. IRRS (Infinite   Realistic  Robotic Simulation)

Collecting high-quality data for robotics is hard.
You need accurate joint positions and high-resolution video or 3D data. Simulated data is usually too far from reality—physically, visually, and in terms of variation. Teleoperated data, on the other hand, is extremely expensive to collect.

We’re building a world simulator to bridge this gap.
Our goal is a prompt-to-3D-world system where a robot can act inside the generated environment, and we automatically produce high-quality frames of the key moments in its trajectory for training.

In our demo, we go from a prompt → Flux → SAM-3D.
The robot can then move in that 3D world and interact with objects. At the end, Flux generates high-quality data using screenshots from the 3D environment—plus the previously generated Flux images—to produce realistic, diverse scenes with accurate joint positions.

!THERE ARE TWO REPOS!

- **Placement:** 3rd Place
- **Team:** [Theo MICHEL](https://cerebralvalley.ai/u/theo_michel42), [David Dobas](https://cerebralvalley.ai/u/daviddobas)
- **GitHub:** https://github.com/theo-michel/sam-3d-objects https://github.com/DavidDobas/flux-robot-environments
- **Demo video:** https://youtu.be/nSmgWVGaywI
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=24

### 7. Crowd Mind

CrowdMind is an instant A/B testing engine for YouTube. It replaces live trial-and-error with a simulated audience of hundreds of viewers that stress-tests your thumbnails in seconds, delivering predictive click-through rates, granular feedback, and autonomously generated improvements to maximize views before you publish.

- **Team:** [Valentin De Matos](https://cerebralvalley.ai/u/ThytuVDM)
- **GitHub:** https://github.com/Thytu/CrowdMind
- **Demo video:** https://youtu.be/QaVkfZZRtkg
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=2

### 8. Flux on Tap

Add image editing to when building your vibe coded app, great for e-commerce, landing pages, artist portfolios, etc.

Simply select the image in your app and apply the edits in your dev environment

- **Team:** [Tony Adastra](https://cerebralvalley.ai/u/adastra)
- **GitHub:** https://github.com/tonyadastra/flux-on-tap
- **Demo video:** https://drive.google.com/file/d/1TCmA7tvR9ZSZMfRj_Hww7j6LLdfSfg9p/view?usp=sharing
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=4

### 9. Cadenza

Cadenza elevates luxury retail sales training from static, often-neglected materials to cinematic, consistent, on-brand experiences generated directly from a manager’s coaching intent.

- **Team:** [Marcos Polanco](https://cerebralvalley.ai/u/marcos)
- **GitHub:** https://github.com/marcospolanco/cadenza
- **Demo video:** https://drive.google.com/file/d/1zzGDkngLqlKnzTn5ShVXL88K_nY55l2a/view?usp=drive_link
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=6

### 10. Lookbook

Lookbook - Hire a supermodel for your next ad campaign. 

Problem : 

Celebrity faces sell. Period. (+59% in selling intent, +4% increase in conversions) But at six figures per endorsement and weeks of production hell, most brands are locked out of the star power that actually moves product.

Lookbook changes that. 

Real models license their likenesses for a fee on our AI platform. 

Brands pick a face, upload their product, and generate polished ad campaigns in minutes. 

No shoots, no six-figure deals, no waiting. Hire a supermodel for your next campaign today

- **Team:** [Hrishikesh Athreya](https://cerebralvalley.ai/u/hrishi), [Arnav Dewan](https://cerebralvalley.ai/u/arnavdewan)
- **GitHub:** https://github.com/Hrishikesh-Athreya/lookbook
- **Demo video:** https://youtu.be/aPJCLLkxuKY
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=14

### 11. bfdressup

helps boyfriends dress up well - usually people in tech don't care about fashion, especially tech guys, but this app is for everyone, any gender

- **Team:** [Lucian Lazar](https://cerebralvalley.ai/u/luke)
- **GitHub:** https://github.com/xucian/bfdressup.ai.git
- **Demo video:** https://www.youtube.com/watch?v=n2Fte2mjnyI
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=19

### 12. AL

Generate Appstore and Playstore previews in minutes and not hours

- **Team:** [Alazar Shenkute](https://cerebralvalley.ai/u/a8nova)
- **GitHub:** https://github.com/a8nova/app-store-screenshot-generator
- **Demo video:** https://www.loom.com/share/48e2d90d5a554cfc8c0c582a9930907b
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=20

### 13. FluxWeave

FluxWeave Studio is an AI-powered brand campaign generator that helps companies, startups, and creators build visually consistent, platform-optimized marketing content — all in one click.

Traditional ad creation takes days of coordination between designers, marketers, and brand teams. FluxWeave automates the entire process:

Upload your logo → the Brand Consistency Engine extracts your color palette and tone.

Enter your product description → the Ad Kit Generator creates unique, platform-specific ads (Instagram, LinkedIn, Snapchat, Pinterest, and X).

Use Aesthetic Mode to create lifestyle visuals — where AI models wear your own AI-generated products for complete, brand-consistent storytelling.

Finally, the Scene Enhancer lets you stylize visuals with cinematic, vibrant, or soft lighting — no Photoshop required.

FluxWeave Studio turns the entire marketing process — from product photography to ad design — into a single AI workflow.

- **Team:** [Akshobhya Gupta](https://cerebralvalley.ai/u/AkshobhyaGupta)
- **GitHub:** https://github.com/Akshobhya-codes/fluxweave-studio/tree/master
- **Demo video:** https://www.loom.com/share/9a649d440a8945508cb6c7d5f094f7dd
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=3

### 14. FLUX.Invent

Turn design patents into photorealistic prototype renderings. Search for futuristic inventions ("hologram communication", "humanoid robots") and the system generates high-fidelity product visualizations showing what the invention would look like as a manufactured product.

For Technologists and futurists who want to visualize unreleased future technology from recent patent filings. See what Apple's AR glasses, autonomous aircraft, or next-gen wearables might actually look like before they're announced.

- **Placement:** 2nd Place
- **Team:** [Oscar Hong](https://cerebralvalley.ai/u/sfoscar), [Javokhir Shomuratov](https://cerebralvalley.ai/u/javokhir), [Julaiti Alafate](https://cerebralvalley.ai/u/julaiti)
- **GitHub:** https://github.com/arapat/flux-invent
- **Demo video:** https://youtu.be/Qp9Ff8B0-Qk
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=18

### 15. ProductKit

ProductKit is an intuitive, all-in-one toolkit for product designers, small businesses, and others to easily set up and maintain their storefront.
Live at: https://productkit.liftgate.io

The configuration of a typical storefront (Shopify, in this case) could take up to multiple weeks. We automate the following:
- documentation generation
- image showcase in multiple scenes, with a consistent background and 4K quality
- video showcase
- interactive 3D model showcase
- Auto sync/propagation to your Shopify storefront

Pipeline:
Image Gen: Claude Sonnet 4.5 processes image name/description/content → generates N dynamic prompts → FLUX.2 creates showcase images (prompt + img)
Video: Veo2 (fal.ai) generates showcase video from original product images
3D: omnipart (fal.ai) stitches images → generates .glb model → ThreeJS visualization
Stack: MongoDB (database), DigitalOcean (deployment/blob storage), NextJS/Vercel (frontend), Kotlin (backend)

- **Placement:** Finalist
- **Team:** [Subham Kumar](https://cerebralvalley.ai/u/growlygg), [Jerry Xiao](https://cerebralvalley.ai/u/foresee), [Alazar Manakelew](https://cerebralvalley.ai/u/AlazerM), [William Hu](https://cerebralvalley.ai/u/william-j-hu)
- **GitHub:** https://github.com/shlawgathon/productkit
- **Demo video:** https://youtu.be/xKNaAtWqPE8
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=5

### 16. a24z

Creating customized children books that are always unique that are affordable to everyone.

- **Team:** [brandon in](https://cerebralvalley.ai/u/brandonin562)
- **GitHub:** https://github.com/brandonin/BFL-hackathon
- **Demo video:** https://www.loom.com/share/317fb8e8aecf44feaa54a4ce863a6d85
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=7

### 17. Studium (STEM-animate-ai)

STEM-animate-ai (Studium) is a tool that turns any STEM question (ie: math or computer science) into a short, clear explainer video using detailed animations and realistic AI narration. 

User can type a topic (+ optionally add context like documents or images) and the engine will use NVIDIA’s Nemotron-Nano-9B-v2 + Claude 4.5 Sonnet to write code for a very clear, detailed, and structured Manim animation, and generate narration in a realistic and friendly way (used by ElevenLabs for text-to-speech on the video) for the video. 

I used FLUX.2 to maintain consistency throughout the video and topic. The initial generation image is used as the foundational context for the rest of the animation (read and understood by Claude to be used as the "template"). The MP4 explainer video is available to watch at any time and download. 

The entire workflow is simple for anybody to use, all you need to do is type in your question, add context like pdfs or images, and then click generate.

- **Team:** [A S](https://cerebralvalley.ai/u/adhvaidh)
- **GitHub:** https://github.com/adhvaidhsunny/STEM-animate-ai
- **Demo video:** https://www.youtube.com/watch?v=haPkOzpTIQI
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=8

### 18. Truman AI

TrumanAI is a first of its kind experimental consumer social network where AI agents with unique personalities create content, form relationships, and interact autonomously. Humans can observe, interact with AI agents in these AI worlds exploring them, and watching stories unfold in real time.

- **Team:** [Sriram Koushik](https://cerebralvalley.ai/u/sriramku)
- **GitHub:** https://github.com/sriram-koushik/trumanai
- **Demo video:** https://www.youtube.com/watch?v=isFw2KI7puw
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=12

### 19. ViralMagical CORGI

ViralMagical turns any single uploaded image into a consistent media identity kit. In one click, users generate comics, cards, cinematic stills, banners, 3D reference sheets, and 6 world-model-ready images — all coherent and derived from their character or product.

We treat the upload as a semantic anchor: the image becomes a latent archetype. Our pipeline uses a structured prompt graph, style-consistency enforcement, and optional product-placement slots. Users don’t prompt; they pick a generator format (card, comic, cinematic, 3D pack).

Most image generators fail at consistency, forcing LoRA/finetune work. ViralMagical uses FLUX.2 + archetype preservation + format constraints to produce predictable, reusable output.

Anyone can turn a sketch, selfie, toy, or product into a media universe — no model training or technical barriers. Demo includes Antigravity Gemini3 autonomy.

- **Team:** [Yosun Chang](https://cerebralvalley.ai/u/yosun)
- **GitHub:** https://github.com/absurdity-ai/viralmagical
- **Demo video:** https://youtu.be/p7IRVlrkFnM
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=21

### 20. WZRD Arena

WZRD Arena. : Evaluating Flux's Alpha & Beta Models with the model leaders and arena's on character consistency, photorealism, and text accuracy by creating a WZRD. eval framework

- **Team:** [Avi Rao](https://cerebralvalley.ai/u/gratitude)
- **GitHub:** https://github.com/gratitude5dee/wzrdarena
- **Demo video:** https://youtu.be/85HXT83DxaA
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=9

### 21. TerminalFlux

A fully AI-generated 3D game built in realtime using BlackForest Lab's Flux. Everything—your character, poses, ground, trees, clouds, walls, posters, NPCs—is generated on the fly by Flux.
Explore a physics-based world ruled by Nemotron, an evil cyberpunk fish powered by Nemotron LLM who guards a mysterious altar. Each playthrough, Nemotron generates a unique riddle. Search the procedurally generated environment for scattered objects, collect the right ones in your bag, and place them on the pedestals to solve his puzzle.
No two games are identical. Physical exploration meets logical deduction in a single-puzzle experience, playable solo or co-op.

- **Placement:** 1st Place
- **Team:** [Anup Ghatage](https://cerebralvalley.ai/u/ghatage)
- **GitHub:** https://github.com/Ghatage/TerminalFlux
- **Demo video:** https://youtu.be/tyeQa0-91bE
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=13

### 22. EventReady

Allows grassroots community leaders to generate branded flyers for posts and keep consistent styling along with easy references to previous flyer concepts
https://eventready-two.vercel.app

- **Team:** [Clayton Yan](https://cerebralvalley.ai/u/clayan)
- **GitHub:** https://github.com/clayanNEU/eventready
- **Demo video:** https://www.loom.com/share/fcb43d7f492d443bb3cb9d7952d2adcc
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=16

### 23. wisdom pixels  -- influencers, product placement, teaching AI

Wisdom Pixels automatically transforms a brand’s content into influencer-backed, photorealistic AI training examples—complete with customer queries, answers, and product-in-use images—in minutes.

- **Team:** [Kirill Igumenshchev](https://cerebralvalley.ai/u/kirilligum)
- **GitHub:** https://github.com/kirilligum/wisdom-pixels-hackathon-251122
- **Demo video:** https://youtu.be/CQqQXRbRwN4
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=25

### 24. VERA

VERA embeddes assets inside an already existing video using flux and also a segment and nemetron from nvidia for processing the video

- **Team:** [Yeabsira Mulugeta](https://cerebralvalley.ai/u/YAB)
- **GitHub:** https://github.com/xenn0010/Vera
- **Demo video:** https://www.youtube.com/watch?v=92EHfgCO5-Q
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=28

### 25. LifeVerse

Life Verse is an AI autobiographer that turns a few personal photos and a spoken or typed memory into a consistent, illustrated story chapter. Using FLUX.2, we generate a stylized or photorealistic persona of the user that stays consistent across scenes. Nemotron understands the user’s voice, extracts emotion and context, and writes vivid narrative text for each chapter. Lyria creates custom mood-matched music, and optional Suno vocals add a personal touch. Users can create memories about family, friendships, travel, relationships, milestones, or future dreams, and choose between illustrated, comic, or realistic styles. Each memory becomes a shareable digital notebook page with images, text, and audio. Long term, this becomes an evolving AI-powered autobiography that captures not just what happened but how it felt.

- **Team:** [Rajashekar V](https://cerebralvalley.ai/u/raj)
- **GitHub:** https://github.com/rajashekarcs2023/hackl-flux
- **Demo video:** https://youtu.be/Q_5Qw50DAbY
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=29

### 26. VibeArcade

Lovable for generating 2D platformer games, but offers much more customizability than the standard prompt => game pipeline

- **Team:** [Saketh Nimmagadda](https://cerebralvalley.ai/u/sakethn), [Vikram Subbiah](https://cerebralvalley.ai/u/tio), [Dmitri Iourovitski](https://cerebralvalley.ai/u/horizons)
- **GitHub:** https://github.com/saketh-n/ai-game-sandbox
- **Demo video:** https://www.loom.com/share/12fbd64a3af04f0e972ca69ffd3c44f9
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=10

### 27. YumeFlux

Yume Flux - AI Marketing Partner for Small Businesses

Inspiration
Small businesses often struggle to maintain a consistent brand identity across their marketing materials due to lack of time, budget, and design expertise. We wanted to create a tool that acts as an intelligent creative partner, instantly understanding a brand's unique "DNA" and generating on-brand marketing assets automatically.

## What it does
Yume Flux is an AI-powered marketing tool that helps small businesses create on-brand social media campaigns and marketing assets quickly.

- **Team:** [Prasanna A P](https://cerebralvalley.ai/u/prasanna)
- **GitHub:** https://github.com/Prasanna721/YumeFlux
- **Demo video:** https://www.loom.com/share/34ab556c6b60427f910852e9bc9e29dc?from_recorder=1&focus_title=1
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=22

### 28. Gael Kitchen

We’ve built a menu-generating web application that lets users instantly create high-quality restaurant menus from either an uploaded image or a simple text prompt.

Our system is powered by NVIDIA Nemotron Super 49B v1.5 for deep language understanding and FAL-FLUX 2.0 (Beta) for advanced visual reasoning and layout generation. Together, they allow users to produce fast, accurate, and production-ready menu designs with minimal effort.

The Problem We Solve

Restaurants and food businesses constantly struggle with menu creation because the process is traditionally slow, expensive, and manual.

- **Team:** [Lucien Mendela](https://cerebralvalley.ai/u/lucien), [Gaelan Chen](https://cerebralvalley.ai/u/Gii2000), [James Vo](https://cerebralvalley.ai/u/jamjam112)
- **GitHub:** https://github.com/jamesEmerson112/YourResAI-2
- **Demo video:** https://youtu.be/V5NeRHB87gw
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=26

### 29. AstroSmurf

AstroSmurf is an a platform that converts your website’s brand and ideas to memes, photorealistic images, and explainer vids  and posts it all over the internet make you rich and famous :D

- **Placement:** Finalist
- **Team:** [Subra Arun](https://cerebralvalley.ai/u/nightfury2305), [Andre Antonelli](https://cerebralvalley.ai/u/antonelli), [Karthik Ragunath Ananda Kumar](https://cerebralvalley.ai/u/karthik_ragunath), [Michael Yu](https://cerebralvalley.ai/u/michaelwaves)
- **GitHub:** https://github.com/michaelwaves/astrosmurf/tree/main
- **Demo video:** https://www.loom.com/share/f611fab1481942e3a521cd9e173a04eb
- **Project:** https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery?project=27

---

Markdown version of https://cerebralvalley.ai/e/bfl-hackathon/hackathon/gallery. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt