Slack Triggers in AACFlow: Build Real-Time Slack-Driven Workflows

• message.created — any new message posted to a channel or DM
• app_mention — messages that directly mention your Slack App (@YourBot)
• slash_commands — custom /commands your team registers in the Slack App manifest
• reaction_added — emoji reactions, ideal for approval workflows (👍 = approve, ❌ = reject)

Each event exposes structured outputs: channel, user, text, ts (timestamp), thread_ts (thread context), team_id, and the raw event payload for advanced use cases.

How to Set Up Slack Triggers in AACFlow

Setting up the Slack trigger takes about five minutes. Here is the exact sequence.

Step 1 — Create a Slack App. Go to api.slack.com/apps, create a new app from scratch, and enable the Events API. Under OAuth & Permissions, add the bot token scopes your workflow needs: channels:read, chat:write, commands, and reactions:read are the most common.

Step 2 — Connect credentials in AACFlow. Open your workspace settings, navigate to Credentials, and add a new Slack credential. AACFlow walks you through the OAuth flow — paste your Slack App's Client ID and Client Secret, complete the OAuth handshake, and AACFlow securely stores the bot token.

Step 3 — Add the Slack trigger to a workflow. In the AACFlow canvas, drag the Slack Trigger block onto the canvas. Select your credential, choose webhook or poll mode, and pick the events to subscribe to. AACFlow generates the webhook URL automatically — copy it into the Slack App's Event Subscriptions page.

Step 4 — Subscribe to events. In the Slack App dashboard, paste the AACFlow webhook URL under Event Subscriptions and subscribe to the bot events that match your trigger configuration. Save the manifest and reinstall the app to your workspace.

Your trigger is live. Send a message in Slack and watch the workflow fire in the AACFlow execution log.

Building a Real-Time Slack Support Bot

Here is a complete workflow that routes incoming Slack support messages through an AI agent and sends an automated reply — all in under three seconds.

Workflow structure:

1. Slack Trigger — listens on message.created in the #support channel, webhook mode
2. Condition block — skips bot messages (user.is_bot === true) to prevent loops
3. AI Agent block — uses claude-sonnet-4-6 with a system prompt that instructs it to classify the request (billing, technical, feature request) and draft a reply
4. Router block — reads the agent's classification output and routes to one of three branches
5. Slack — Send Message — posts the reply to the original thread using thread_ts, keeping conversations organized

The AI Agent block in step 3 receives the full Slack event context: the message text, the channel name, the user ID, and the last 10 messages from the channel (fetched via the Slack — Read Channel block). This context lets claude-sonnet-4-6 draft replies that reference the conversation history rather than treating each message in isolation.

The Router block is one of AACFlow's most powerful primitives for Slack workflows. You define route descriptions in plain language — "billing questions about invoices and payments", "technical bugs and errors", "feature requests and suggestions" — and the router uses an LLM to classify the incoming message against those descriptions. No regex, no keyword lists.

Advanced Slack Trigger Patterns

PR notification workflows. Combine the GitHub webhook trigger with the Slack — Send Message block to post richly formatted PR summaries to the #engineering channel. Use the channel-selector subblock in the Slack block to let team members configure the target channel without touching the workflow canvas.

Incident alert routing. A PagerDuty or Datadog webhook fires into AACFlow, which uses a Router block to determine severity, then posts to #incidents (P1/P2) or #alerts (P3/P4) via the Slack block. The reaction_added trigger on the incident message starts the acknowledgment flow when an on-call engineer adds the 👀 emoji.

Async standups. A Schedule trigger fires every morning at 9:00 AM per timezone. AACFlow uses the Slack — Send Message block with the slash_commands mode to prompt each team member with three standup questions. Replies collected via app_mention are aggregated by an AI Agent and posted as a summary to #standups.

Multi-channel routing. When a single AACFlow workflow serves multiple Slack channels, the Router block inspects the channel output from the trigger and dispatches to channel-specific logic branches. The Slack — Read Channel block in each branch fetches channel-specific context before the AI Agent responds.

AACFlow's Slack integration is part of its 304+ connector catalog. Slack workflows can call into any other connector mid-flow — look up a customer in HubSpot, create a Jira ticket, query a PostgreSQL database, or trigger a child workflow — all without leaving the Slack conversation context.

Conclusion

AACFlow's Slack trigger — with webhook and poll modes, four event types, and deep integration with the AI Agent and Router blocks — gives teams a production-grade path from Slack events to automated outcomes. The five-minute OAuth setup, structured event outputs, and native Slack block operations (Send Message, Read Channel, channel-selector) mean you spend time designing workflow logic, not wiring up API plumbing.

Start with the support bot pattern, then extend it with the Router block to handle multi-intent messages. Once you see claude-sonnet-4-6 classifying and responding to Slack messages in real time, the next use case — incident alerts, PR notifications, async standups — becomes obvious.

Try the Slack trigger in AACFlow today and connect your team's most-used communication layer to the full power of AI-driven automation.