> For the complete documentation index, see [llms.txt](https://notara-1.gitbook.io/notara-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://notara-1.gitbook.io/notara-docs/getting-started/channels.md).

# Inviting the bot to channels

The Notara agent can be active in any number of Slack channels. Each channel can have its own tool set, context configuration, and permission overrides — so your engineering channel can have different tools than your marketing channel.

## Inviting the Bot

### From Slack

In any Slack channel, type:

```
/invite @Notara
```

Replace `Notara` with your bot's configured display name. The bot will join the channel and confirm with a brief greeting.

### From the Dashboard

1. Go to **Connect → Channels** in the Notara dashboard.
2. Click **+ Add Channel**.
3. Select the Slack channel from the dropdown.
4. Configure the channel settings (see below).
5. Click **Save**.

## Configuring a Channel

After adding a channel, click its name in the channel list to open the configuration panel. Here's what you can configure:

### Tool Profile

Choose a preset tool profile for this channel:

| Profile         | Tools Included                                   |
| --------------- | ------------------------------------------------ |
| **Full**        | All connected tools                              |
| **Engineering** | GitHub, Linear, Slack, Memory                    |
| **Growth**      | Stripe, Perplexity, Firecrawl, ChartMogul, Slack |

You can also customize the tool set by toggling individual tools on or off from the **Tools** tab within the channel config.

### Context Collections

Select which context libraries are available in this channel. The agent will have access to the documents in these collections, respecting each document's load policy (spine vs on-demand vs never).

### Tool Permission Overrides

Override the workspace-level tool permissions for this specific channel. For example, you might allow `write` tools without approval in a private engineering channel, while requiring approval in a general channel.

Permission overrides follow the same four-layer resolution order:

```
capability default → tenant permissions → channel override → active skill
```

See [Tool Permissions](/notara-docs/team/permissions.md) for details.

### Active Skill Profile

Bind a skill (or set of skills) that are active by default in this channel. When a user's message matches a skill's intent, the agent automatically invokes it.

## DM vs Channel Interactions

The agent behaves slightly differently in DMs versus channels:

| Behavior         | Slack Channel                                             | Direct Message                |
| ---------------- | --------------------------------------------------------- | ----------------------------- |
| Trigger          | Mention `@Notara` or reply in a thread started by the bot | Any message                   |
| Context          | Channel's configured context and tools                    | Workspace defaults            |
| Thread replies   | Always replies in-thread                                  | Continues the DM conversation |
| Approval prompts | Sent to admin's DM                                        | Sent to admin's DM            |

**In channels:** the agent only responds when explicitly mentioned (`@Notara`) or when replying in an existing bot thread. This prevents it from noisily intercepting every message.

**In DMs:** the agent treats every message as a request. This is the best place for one-off queries and personal productivity tasks.

## Removing a Channel

To remove Notara from a channel:

1. In the dashboard, go to **Connect → Channels**.
2. Click the channel, then click **Remove Channel**.
3. In Slack, you can optionally type `/kick @Notara` to remove the bot from the channel entirely.

Removing a channel configuration does not delete any context collections or skills — only the channel's configuration is removed.

## Multiple Channels, One Workspace

There's no limit to the number of channels you can configure. Common patterns:

* **`#ai-help`** — a dedicated channel for AI-assisted work, full tool access
* **`#engineering`** — engineering-specific tools (GitHub, Linear), engineering library
* **`#growth`** — growth tools (Stripe, Perplexity, ChartMogul), competitor research library
* **`#leadership`** — minimal tools, high-trust context (financials, strategy docs)

Each channel maintains independent context and tool configuration, but all channels share the same workspace, model key and team membership.