Agent setup guide
Customer Support Agent
Triages every support message, escalates urgent ones to Slack and drafts replies for approval.
What it does
Triages every support message, escalates urgent ones to Slack and drafts replies for approval.
- Receives each new support message from your contact form, chat widget or helpdesk through a webhook.
- AI tags the category, urgency and sentiment, writes a two-line summary and drafts a reply using only the policies you paste in.
- Urgent or angry messages are posted to your Slack support channel with the summary and suggested reply, so a person answers right away.
- Every other message waits in Reviews. When a teammate approves the draft, it's emailed to the customer with Reply-to set to your support inbox.
- Every message and outcome is kept in the run history and in Agent analytics.
What it needs
- A paid plan (agents run live on paid plans).
- An AI provider key (Settings → AI Providers).
- Slack, connected in Settings → Integrations.
- A form, chat widget or helpdesk that can send new messages to a webhook URL.
What it doesn't do
- It doesn't read an email inbox or WhatsApp by itself; messages arrive through the webhook.
- It never emails a customer without a person approving the reply first.
- It doesn't look up orders or accounts, or issue refunds. It says a teammate will check instead.
- It doesn't update tickets in your helpdesk.
Availability and plans
Available
Its workflow is built and tested, and every app it needs is live. You can get it and set it up now.
Included free on the Pro, Business, Elite and Enterprise plans; on other paid plans it's a one-time purchase per workspace (the price is on the Agents page). The Agents page in the app always shows its current status.
Before you start
- A paid plan: agents run live, and the Free plan only runs Dry Runs.
- An AI provider key (OpenAI, Anthropic, Google and more) in Settings → AI Providers. AI usage is billed by your provider.
- These apps, connected in Settings → Integrations: Slack.
- A form, app or system that can send each new item to a webhook URL (the agent starts from a webhook).
- About 15 minutes.
Set it up step by step
Get the agent first (buy it, or it's already yours when your plan includes it), then in Agents → My agents press “Set up this agent”. That creates the agent's workflow in your workspace (it counts towards your plan's workflow limit) and opens the canvas with a setup checklist above it. Each item links to the exact place to go and ticks itself off when it's done.
1. Connect your AI provider
The AI steps run on your own key (OpenAI, Anthropic, Google and more). Add one in Settings → AI Providers.
2. Connect Slack
Connect the Slack workspace your support team uses (Settings → Integrations → Slack).
3. Pick the Slack account and channel
Open “Alert the team in Slack” and pick your account and the channel for escalations. Invite the app to private channels first (/invite @Fluxo Flow).
| Setting | What to enter | In step |
|---|---|---|
| Slack account | The Slack workspace the alert is posted to. Pick your connected Slack account. | “Alert the team in Slack” |
| Slack channel for escalations | Pick the channel your support team watches, e.g. #support. | “Alert the team in Slack” |
4. Paste your support policies
Open “Triage and draft a reply” and replace the placeholder with your policies and team name.
| Setting | What to enter | In step |
|---|---|---|
| Your support policies | Replace the placeholder in the AI step's instructions with your refund, shipping, billing and contact policies and your team name. The agent only states facts written there. It starts as “[PASTE YOUR SUPPORT POLICIES HERE]” until you replace it. | “Triage and draft a reply” |
5. Add your support email address
Open “Email the reply to the customer” and enter the address customers should reply to.
| Setting | What to enter | In step |
|---|---|---|
| Your support email address | Customers' answers to the reply come back to this address (Reply-to). | “Email the reply to the customer” |
6. Test it with Dry Run
Press Dry Run: the agent runs on sample data with no messages sent and nothing written to your apps.
7. Switch the trigger on
Open the trigger, switch the webhook on and copy its URL into your form, chat widget or helpdesk.
How it works
The agent is a normal workflow with 8 steps (1 AI step, with a person approving in Reviews). You can open, edit and monitor it like any other workflow.
- New support message — Starts when your form or app sends data to the agent's webhook URL. Your contact form, chat widget or helpdesk posts each new customer message to the workflow's webhook URL.
- Triage and draft a reply — AI step (runs on your AI provider key). AI tags the category, urgency and sentiment, writes a two-line summary and drafts a reply grounded only in your policies.
- Urgent or angry? — Decision: picks the branch to follow. Urgent or angry messages go straight to your team in Slack; everything else gets the drafted reply for approval.
- Alert the team in Slack — Slack: Post a message. Branch: Urgent or angry. Posts the customer, category, summary and the suggested reply to your support channel so a person answers it now.
- Escalated — Kept in the run history. The run history keeps the triaged message for your records.
- Approve the reply — A person approves or rejects it in Reviews (the run waits). Branch: Normal. A teammate reads the drafted reply in Reviews, edits it if needed and approves it, or rejects it to answer personally.
- Email the reply to the customer — Sends an email from your workspace's sender. Branch: Approved. Sends the approved reply to the customer from your workspace's sender, with replies going to your support inbox.
- Answered — Kept in the run history. The run history keeps the message and the reply that was sent.
Test it with Dry Run
Press Dry Run in the canvas. The agent runs on the sample data stored on its trigger (the “Sample data for Dry Run” box), so you can check every step's output before anything is live. Edit the sample to match the fields your own form or app sends.
Sample data fields
name, email, subject, message, source.
- Nothing is sent or written: app actions, emails and Slack posts are simulated, and AI steps don't call your AI provider.
- Approval steps use the Dry Run decision you pick in the step's settings (approved, rejected or timeout), so you can try each branch.
Go live
- Open the trigger, switch the webhook on and copy its URL into your form, chat widget or helpdesk.
- The webhook is protected by a secret: send it in the header shown in the Activation panel (requests without it are refused with 403). Each request starts one run.
- The setup checklist adds a “paid plan” item before activation: live runs need a paid plan (the Free plan only runs Dry Runs).
- When every checklist item is done and the trigger is on, the checklist says “Live — this agent is running”. Follow its runs in the workflow's History tab and in Analytics → Agents.
Questions
Does it email customers without a person checking?
No. Every reply waits in Reviews until a teammate approves it. Urgent and angry messages aren't answered automatically at all; they go to Slack.
Can it read my inbox?
Not by itself. Send messages to its webhook from your form, chat widget or helpdesk (most can call a webhook on new tickets).
Troubleshooting
- “No AI provider connected” or a rejected key: connect or fix the key in Settings → AI Providers, then press Test there.
- A checklist item stays unticked: the setting is still empty or still contains its placeholder text. Open the step from the checklist and fill it in.
- An app step fails with “Needs reconnect”: the account's access was revoked or expired. Reconnect it in Settings → Integrations.
- Slack says the app isn't in the channel: in Slack, type /invite @Fluxo Flow in that channel (always needed for private channels).
- Nothing happens when your form sends data: check the webhook is switched on and the request carries the secret header (403 means it's missing or wrong).
- A run shows as paused: it's waiting for someone in Reviews to approve or reject it.
- A step failed: open the run in History and check that step in the Trace tab for the error and what it received.
Generated from the agent's own workflow, so it always matches what the agent does. Last updated .