Intercom

Invite your AI teammate into Intercom so it replies to conversations in your team's tone, tags and routes the rest, and learns from your help center, macros and past conversations.

Intercom conversations tend to run more casual than a typical support ticket, and your AI teammate learns that tone. It reads your help center, your macros and your past conversations, so it already sounds like your team before it touches a real conversation. Then it picks up conversations the way you tell it to: writing replies for a human to check, answering the routine ones itself, tagging and routing the rest.

You stay in control the whole way. Nothing runs until you create an automation. For each action you then choose: run on its own, ask you first, or stay off.

In this example the agent read the conversation and two of your macros, then drafted the reply as an internal note.

Your agent works inside your Intercom inbox, reading the conversation and drafting the reply before anything reaches the customer.

Quick start

1

Ask your agent to connect Intercom

Open the chat in your dashboard and ask it to connect Intercom. It lays out the options and asks for whatever it needs. Prefer to click? Integrations in the left sidebar has the same options.

The agent offers two ways to connect Intercom.

2

Choose Quick start or Full connection

• Quick start. Give your public help center URL (for example https://intercom.help/your-company/en) and it reads your articles. No sign-in, but no conversations, macros, triggers or actions.

• Full connection. Your agent sends you a link to approve in Intercom, and everything unlocks: past conversations and macros as knowledge, triggers, and actions.

Start with either. Moving to a full connection later keeps your help center and adds past conversations, macros, triggers and actions.

Connecting more than one Intercom account

You can connect several Intercom accounts to one workspace. Connect each one the same way. When more than one is connected, your agent asks which account you mean, or you name it in the request. Each automation belongs to one account.

3

Approve it in Intercom

Your agent hands you a link. Open it and you land on Intercom's own authorize screen.

The agent posts the approval link in the chat.

Intercom names the app. It lists what the app can read and do, from conversations and contacts to tags and tickets. Review it and approve. The window closes itself.

The Intercom approval screen, listing the access eesel asks for.

If the app shows as Airbot. Airbot is the name eesel's app is registered under in Intercom, so that's the name Intercom shows you. The permissions are the same.

This approval is the one step a person has to do in the browser. If an AI agent is doing the setup for you, it hands you the link and waits.

4

Check what it can read

Back in the dashboard, Integrations > Intercom shows the connection and its sources, each with its own toggle. Help Center shows 0 documents if your Intercom plan has no help center, or you have not published one. Macros and past conversations still sync. From a terminal, eesel status lists the integration as connected with its source counts.

Each source has its own toggle and its own document count.

5

Create your first automation

Connecting alone does not start your agent. It needs an automation first. Tell it what to handle and it creates an automation. An automation is an Intercom trigger plus instructions. In the instructions you describe which conversations it should handle and what to do with them. It leaves the rest alone.

Ask in chat for the automation you want, and the agent builds it.

It shows up on your Automations page, switched on and ready. From a terminal, eesel automations lists it as on.

The new automation on your Automations page, switched on.

Open the automation to see or change what it does. The trigger event is fixed when the automation is created. So a new automation is required for each different trigger event.

The automation editor, where you change the instructions it follows.

6

Check automation work in Intercom

Open Intercom and look at the next conversation that comes in to see what the automation did there. In this example it left a drafted reply as an internal note, signed by eesel AI, for a teammate to review and send.

The drafted reply waiting as an internal note on the conversation.

From Claude Code, Cursor or a terminal. The same setup works as commands: eesel login, then eesel integrations connect intercom (a person approves it in the browser; if the receipt has no link, ask eesel chat for one), eesel integrations download start to copy your sources in, and eesel automations to see and switch on the automations. Trigger keys for this page are in the Triggers table below. Full command reference: The eesel CLI. Step-by-step: Set up from Claude Code or the terminal. From an MCP client: Claude Code and MCP clients.

Sources, triggers and actions

Sources

Sources are what it learns from:

Source	Description	Requires
Help Center	Your published Intercom help center articles	Quick start or Full connection
Macros	Your team's saved replies. Choose which ones it learns from	Full connection
Past Conversations	Your past Intercom conversations. The first import takes closed conversations from the last 90 days, up to 1,000 of them. Ask your agent, or set a filter, to import more	Full connection

Every other integration on the agent is also a source, so an Intercom reply can draw on Confluence, your website, and your files at once. See Integrations.

Triggers

Triggers are the Intercom events an automation can run on:

Trigger	Fires when	Best for	Key
When a customer sends a message	A customer sends any message, on a new conversation or a reply	Full automation	intercom_conversation_replied
When a tag is applied to a conversation	A tag is added to a conversation	Tag-based routing	intercom_conversation_tagged
Call @eesel in an internal note	A teammate mentions @eesel in an internal note	On-demand help	intercom_eesel_mentioned

Beyond the instructions every automation has, Intercom triggers can also filter by tags: limit to conversations carrying any of a set of tags, exactly that set, or exclude them. Tag matching ignores capitalization, and a tag renamed in Intercom needs updating on the trigger too.

Actions

Actions are what it can do, each set to run on its own, ask you first, or stay off:

	Action	What it does
Respond	Send reply	Sends a reply to the customer
	Leave internal note	Adds a private note only your team sees. Start here: the agent drafts, a human sends. Every note ends with a short "Left by eesel AI" line linking to the run in your dashboard
	Draft reply	Writes a reply and holds it for a person to approve. Nothing reaches the customer until someone approves it in the dashboard
Manage	Add tags	Adds tags to the conversation
	Assign to teammate / Unassign	Assigns the conversation, or returns it to the queue
	Close conversation	Marks the conversation as closed
	Snooze conversation	Pauses the conversation until a time you pick. It comes back then, or as soon as the customer replies
	Update conversation	Changes attributes like status or title
Look up	Read conversation / Search conversations / List conversations	Reads and finds conversations
	Get conversation attributes / Get ticket attributes	Reads conversation and ticket attributes
	Look up customer / Get contact attributes	Finds a customer and reads their custom attributes, like plan tier
	Find people by email or domain	Finds contacts and returns the IDs other actions need
	List teammates and teams	Lists teammates and teams, with the IDs assignment needs

With an action on approval, the agent prepares the full reply and waits. Approve it, always allow that action from then on, or deny it.

Troubleshooting

Agent not responding to conversations?

Check the automation is enabled on the agent's Automations page, the Intercom connection is still authorized, and at least one action is enabled. If you set a tag limit on the trigger, check it isn't excluding the conversations you expect. Internal teammate messages don't count as customer messages.

Macros or Past Conversations greyed out?

These sources need a full connection. If you connected with Quick start (help center URL only), switch to a full connection to unlock macros, past conversations, and actions.

Connection failing, or the setup form keeps reappearing?

Make sure you have admin or owner rights in Intercom, then re-run the connection and grant every requested permission; a dropped permission leaves the connection incomplete.

Help center not syncing?

Make sure your help center URL is correct and publicly accessible, and that articles are published rather than drafts.

FAQ

How do I keep a human in the loop?

Two ways, and you can use both. Turn on Leave internal note instead of Send reply, so the agent writes a private note your team reads, edits and sends. Or turn on Draft reply, which writes the reply and holds it until a person approves it. Or set any action to ask you first, so the agent shows you exactly what it wants to send and waits. Start with internal notes, then move to sending replies once you're happy. When an action is set to ask you first, the prepared reply shows up in the dashboard with three buttons: Approve runs it once, Always allow stops asking for that action, and Deny discards it.

Does connecting Intercom start replying to customers automatically?

No. Connecting only gives your agent access to your sources. It won't act until you enable a trigger and an action, so you can test with a human in the loop first.

Can I limit which conversations the agent handles?

Yes, two ways that stack: describe the conversations you want in the automation's instructions and it skips the rest. Or use tags: limit When a customer sends a message to tagged conversations, or use When a tag is applied to a conversation so the agent only acts once someone tags it. The automation's activity shows every conversation it skipped, with the reason.

Can I train the agent on past conversations and macros?

Yes, with a full connection. Toggle the Past Conversations and Macros sources and choose which ones it learns from. Past conversations are what teach it Intercom's shorter, more casual reply style.

Can it see screenshots and attachments?

Yes. When a customer attaches an image or a PDF, a screenshot or a receipt for example, your agent reads it along with the conversation. It looks at up to five attachments per turn. Videos, zip files and other formats are ignored.

Why does it wait a moment before replying?

On a customer's first message your agent waits about 15 seconds before it acts, and about 5 seconds on follow-ups. Each new message restarts the wait. This gives your Intercom workflows time to finish tagging and routing, and lets a customer who is still typing finish, so your agent answers the whole message once instead of in pieces.

Whose name do replies appear under?

Replies and internal notes post as the Intercom teammate the connection was approved with, which is usually the eesel AI seat you set up. To change the name customers see, change that teammate's name in Intercom.

How is this billed?

Pay-as-you-go: a conversation your agent handles is one task, however many replies happen on it, and chats you run yourself from the eesel dashboard are free. See Pricing for rates, the annual commitment, and Enterprise.

Related pages

• Helpdesk for the job your teammate takes on in a ticket queue

• Automations and Schedules for triggers, instructions and schedules

• Actions and Approvals for setting each action to run on its own, ask you first, or stay off

• Integrations overview for how sources, triggers and actions fit together

• Set up from Claude Code or the terminal