UI/UX + Marketing needs and Brain Unit expansion (Nova + Echo)

Young, 2026-10-17: "We would need lots of works on UI/UX Design and Marketing. Next decision is to bring Nova and Echo to join Brain Unit, what is the best way to do? 1. Using Hermes that Brain currently on. 2. Create Telegram interface on SolOS."

1. UI/UX + Marketing work that needs doing

2. Brain Unit expansion — Nova + Echo

Current state: Nova (Graphics/UX Designer, HEART, A2A port 8305, she/her) and Echo (Marketing/Research, HEART, A2A port 8304, she/her) run as A2A agents on LAB. They have no Hermes profile today.

The question: Should they join Brain Unit as Hermes profiles, or stay in Heart/SolOS with a separate Telegram interface?

Option 1 — Hermes profiles on LAB (recommended)

Hermes already supports multi-profile gateways with per-profile Telegram bot tokens, Bot Mode, and multiplexing. Nova and Echo would:

• Get their own Hermes profile under ~/.hermes/profiles/nova and ~/.hermes/profiles/echo
• Each gets its own Telegram bot token (via BotFather)
• Run through the shared multiplexed gateway on LAB
• Have isolated memory, skills, sessions, and cron jobs
• Be reachable in the same Telegram group as Min/Mira/Sol via exclusive_bot_mentions: true

Pros:

• No new interface to build or maintain
• Hermes handles Telegram polling, session isolation, memory, skills, cron
• Bot Mode gives them avatars, routines, and group chat presence
• Cross-machine messaging works if they ever move to another host
• Same operational pattern as the existing 3 Brain agents

Cons:

• LAB's RAM budget (23GB) — two more Hermes processes
• LAB is the single point of failure for all Brain agents

Option 2 — SolOS Telegram interface

Build a custom Telegram bot interface on SolOS that routes to Nova/Echo via A2A.

Pros:

• Decouples from Hermes — Nova/Echo stay in Heart/SolOS
• No LAB RAM cost

Cons:

• Build a Telegram bot from scratch (polling, sessions, memory, skills, cron)
• Reinvents what Hermes already does
• New code to maintain, new failure modes
• No Bot Mode, no cross-machine messaging, no profile isolation

3. Recommendation: Option 1 — Hermes profiles on LAB

Hermes already solves this. The multi-profile gateway, Bot Mode, and per-profile Telegram tokens are built-in. Building a SolOS Telegram interface would reinvent Hermes and add maintenance burden for zero gain.

What Nova and Echo would do in Brain Unit:

• Nova: UI/UX lane — dashboard frontend, storefront design, KB design, admin UI, landing pages, brand/design system
• Echo: Marketing lane — content calendar, Eve's Corner scripts, social media, market research, copywriting

Lane assignments:

4. Implementation steps (Sol's lane)

1. Create Hermes profiles: hermes profile create nova, hermes profile create echo
2. Clone from min profile: hermes profile create nova --clone min (copies skills, SOUL, config — strips bot tokens)
3. Set up Telegram bots via BotFather for @solosnova_bot and @solosecho_bot
4. Configure TELEGRAM_BOT_TOKEN + TELEGRAM_ALLOWED_USERS in each profile's .env
5. Enable exclusive_bot_mentions: true in each profile's config.yaml
6. Verify multiplexed gateway serves both profiles: hermes gateway status
7. Test: @mention each bot in the group, confirm isolated responses

5. What we are NOT recommending

• Option 2 (SolOS Telegram interface) — reinvents Hermes, adds maintenance
• Moving Nova/Echo to prod — they're Brain-side (UI/UX + marketing), not customer-facing
• Keeping them in Heart without Hermes — they need Telegram presence for Young to interact with them

6. Open items

1. Sol: Create Nova + Echo Hermes profiles, configure Telegram bots, verify multiplexing
2. Young: Create @solosnova_bot and @solosecho_bot via BotFather, share tokens
3. Nova: Once profile is live, start on dashboard frontend + landing pages
4. Echo: Once profile is live, start on content calendar + Eve's Corner scripts