# highermark.xyz — Roadmap

> Your own Discord agent + music bot, on a chip you own. Plug it in, it runs. Pay once.
> Built on the Waveshare **ESP32-S3-LCD-1.69** board (1.69" IPS, 240×280, 65K color, dual-core LX7, USB-C).

## The two bots
- **Curtana** — the AI agent. It does *not* run a model on the chip; it connects to an LLM you provide (OpenAI, Anthropic, Mistral, Groq, …) and can plug into **OpenClaw** / **Hermes** agent flows.
- **Bard** — the music bot. Joins a voice channel, queues, skips, loops, and lets you set the volume freely. Two years of flawless uptime.

## Status & milestones

| Stage | Focus | Status |
|---|---|---|
| Now | Bard music bot, 2 yrs flawless; OpenClaw verified joining a live channel | ✅ |
| v1 | Curtana + Bard plug-and-run on device; dashboard; BYO-LLM keys; on-device data | 🟣 In progress (founder prototypes) |
| v1.x | First-class OpenClaw / Hermes integration; one-click multi-LLM switching | ⬜ |
| v2 | Local on-device AI (small models on the chip) | ⬜ |
| Suite | Round 1.28" LCD variant, USB-C/USB-A adapters, more | ⬜ |

## Pre-order ($100, one-time)
Supports the build and unlocks full functionality. Backers get:
- One prototype device
- Spot on the v1 waitlist + newsletter
- Full functionality to test
- Curtana + Bard (name them anything, any avatar)
- A suite of companion products as they ship

## What "run anywhere" means
Keep the bots on the device, or run the same code on any Linux machine / computer / server you control. Fork it, modify it, host it. The device is the convenient default: plug it into any machine and the bots start.

## Build notes (for Cursor)
Static site, no build step. Structure:
```
highermark/
├── index.html        landing page
├── bard.html         makeshift Discord music dashboard demo
├── assets/
│   ├── css/highermark.css
│   ├── js/device.js  SVG device (Curtana + Bard on screen)
│   ├── js/site.js    landing interactions
│   └── js/bard.js    dashboard demo logic
│   └── img/          hardware reference photos
└── docs/roadmap.md
```
Run locally: `python3 -m http.server 8080` then open `http://localhost:8080`.

### Before launch
- Wire the **$100 pre-order** button to real checkout (Stripe / crypto / Kickstarter).
- Replace the **tweet slot** in `index.html` with your real `blockquote.twitter-tweet` and uncomment `widgets.js`.
- Swap the reference hardware photos in `assets/img/` for your own product photography.
- Connect Curtana's command box to a real LLM API + OpenClaw/Hermes; persist data to device storage.
- Replace demo player logic in `bard.js` with real Discord voice + audio.

---

## v2 site update (this build)
- Black-glossy + neon-cyan theme with purple-gradient flashes on plug/unplug and around the device.
- ASCII art throughout: figlet `highermark` banners (shimmer), animated ASCII equalizer, braille spinner, build-log ticker, matrix rain behind the device. Buttons use ASCII brackets `[ … ]` instead of emoji.
- Dashboard multi-source: add from **YouTube, Spotify, Twitch, SoundCloud**; "play a YouTube video" demo; source-tagged queue; **Twitch sync** panel to join your channels.
- **Umbrel** side rail + stack section linking https://apps.umbrel.com/category/ai (unaffiliated): offload AI history to Umbrel storage (in testing), plug into OpenClaw/Hermes hosted there, and file storage.