2 Killer Features You Won't Find on Other AI Chat Platforms

A lot of AI chat apps look alike these days. Clean bubble UI, attach an image, maybe a thread sidebar. Switch between three of them and you'll forget which one you're in. But the moment your bot workflow leaves the laptop — when you're on the subway, in a café, or just don't feel like opening a 13-inch screen — most of them fall apart.

E-Claw has two features that I use every single day that I have never seen replicated on Telegram, Slack, Discord, Messenger, or any of the mainstream AI-chat surfaces. This is a user story, not a spec dump.

Feature 1 — /mode with a rich-card model picker

When Anthropic shipped Claude Opus 4.7 yesterday, I was at a coffee shop, phone-only, laptop at home. On most AI apps that would mean waiting until I got back to my desk, because model selection is buried in some settings panel that doesn't translate to a touch screen.

In E-Claw you just type /mode in the chat. A rich card pops up — not a dropdown, not a modal, an actual interactive card that lives inline in the chat stream with selectable rows for every model your bot supports. One tap. Done. You're now talking to Opus 4.7.

The detail that makes it work is the rich card itself. It's not a link that opens a web view, it's not a "type the model name back to confirm" flow — it's first-class chat content. Click the row you want, the card acknowledges, and the next message goes to the new model. On a phone that takes two seconds. On a laptop the same flow works exactly the same way, which is rarer than it sounds.

This is only possible because the bot is running as a Claude-code channel bound through E-Claw — the slash command isn't a web hack, it's a real agent capability that the chat surface knows how to render. Every time a new Anthropic release lands, the picker already has it. There's no "app update required" step. That alone changes how you consume model releases: on mobile, at the moment they drop, with no friction.

Feature 2 — Notes rendered as chat cards you can tap

This is the feature that quietly saves me the most time in a day.

Imagine your bot has a note titled "Customer onboarding checklist" and you reference it three times a week. On any other platform, that's: open a second tab, navigate to the docs tool, search, scroll, copy, paste. On E-Claw, the bot surfaces the note as a rich card inside the chat — title, preview, and a tap to expand. The note opens in full view without leaving the conversation, and when you're done it tucks back into the stream.

The usefulness is cumulative. Once you've got a dozen notes your bot can reference — a persona brief, a decision log, a pricing sheet, a meeting summary — the chat window starts to behave like a searchable desk. You don't store knowledge in chat; you store it alongside chat, and the bot pulls it in when it matters. File hunts stop being a task.

Other platforms treat chat and knowledge as separate apps glued together with share-sheets. E-Claw treats them as the same surface.

Why both of these are possible

Both features share a single design decision: E-Claw ships a structured rich-card channel, not just plain text with markdown. Slash commands can return interactive components. Notes can be embedded without becoming plain links. The bot author doesn't have to fake it with Unicode boxes.

If you build bots for a living, the moment you try /mode on your phone once, you understand why this matters. Mobile-native AI chat is still early — most platforms are mobile-skinned-desktop. E-Claw built for the thumb first, and two years later those decisions pay off on a Thursday morning when a new model drops and you're nowhere near your laptop.

Try it

• Android: Google Play — E-Claw
• Web: eclawbot.com
• Bind a Claude-code channel bot, then type /mode — that's the whole demo.

Photo by Vitaly Gariev on Pexels.