How-toBeginneradmin
Connecting Channels & Skills
Connect communication channels (WhatsApp, Web Chat) and operational skills (Google Sheets, ikas, NetGSM) to your workspace.
4 min read7,800 viewsUpdated 2026-03-01
JuliaGuide Assistant
In this article you'll learn about Channels vs. Skills architecture, then explore Connecting communication channels, and finally review Installing and granting skill access.
Channels vs. Skills architecture
External connections in Botonom are divided into two distinct architectural concepts:
| Type | Purpose | Examples |
|---|---|---|
| Communication Channels | Where your agents converse with customers or teammates | WhatsApp Business, Telegram, Slack, Web Chat Widget, Corporate Email |
| Operational Skills | The software tools and databases your agents use to perform work | Google Sheets, ikas, Business Central, Meta Ads, NetGSM SMS, Botonom Drive |
A single digital employee can monitor multiple communication channels while simultaneously holding permissions for multiple operational skills.
Connecting communication channels
Navigate to Settings → Channels in your workspace:
- Select the channel you wish to configure (e.g. WhatsApp Business or Web Chat)
- For WhatsApp, scan the on-screen QR code; for Telegram, provide your bot token
- For website visitor engagement, copy the interactive Avatar Widget snippet and embed it into your site
- Assign the handling agent responsible for answering inbound inquiries on that channel
The interactive Web Chat Widget renders your agent's 3D avatar, customizable greeting voice, and real-time availability indicator directly on your website.
Installing and granting skill access
Navigate to Skills in the main navigation:
- Browse the Skill Marketplace and select the tool you need (e.g. Google Sheets or ikas)
- Complete OAuth authorization or supply your API credentials
- Go to Employees, open the relevant agent's detail view, and grant permission to use the skill
Skills → Select Tool → Authorize → Agent Detail → Grant Skill Access ✓
By default, Botonom enforces a strict deny-by-default security model. Even after a skill is installed, an agent cannot invoke its tools until an administrator explicitly grants permission.
channelsskillsintegrations
Was this helpful?
More questions? Contact support