Trengo logo Trengo automation · runs daily

Triage new Trengo conversations with AI

New conversations land in Inbox > New from email and WhatsApp overnight, and every morning someone reads each one to choose a label and a person.

Cotlet does it in your own Trengo account with list_labels, list_assignees, list_conversations, get_conversation, add_label, assign_conversation and add_internal_note, called by your AI agent, an n8n workflow or a script.

Hosted · nothing to install or maintain

Can Trengo do this on its own?

Trengo triages by fixed rules and by its own AI (checked 20 Oct 2026). A rule matches text and contact conditions, then attaches a label or assigns. AI Labeling in Journeys tags closed conversations from the label group descriptions you write, and the AI Agent applies labels and internal comments only while it handles a conversation. None of them reads a new conversation, picks a label and an assignee by your rule and leaves the reason in a note.

How you do it in Trengo today

  1. In Inbox > New, an agent opens each new ticket and reads it.
  2. The agent clicks the label icon and picks a label from the list set up under Settings > Settings > Labels; the label then shows at the top of the ticket and in the inbox summary.
  3. Next the agent clicks Assign and picks a user or a team, or Assign to me to take the ticket.
  4. Repeating cases go into rules under Settings > Automation > Rules, with conditions such as Message body contains "order" or Contact email contains "@gmail.com", joined by AND or OR, and actions such as Attach label or Assign to user.
  5. Bulk actions in the inbox close, assign, mark as spam or delete up to 50 conversations, but labels are not on that list, so labels go on one by one.

How to triage new Trengo conversations with AI

Prompt

Go through the new Trengo conversations. Label each one by topic (Billing, Delivery or Bug), assign Billing to the Billing team, Delivery to Sam and Bug to the Tech team, and leave an internal note with the reason. Leave unclear ones for me.
  1. 01

    Learn the labels and people

    The agent calls list_labels and list_assignees once per run and uses only labels that exist and are not archived, and users and teams as Trengo lists them. A topic with no matching label is reported, never invented.

  2. 02

    Read the new queue

    list_conversations with folder new brings the queue, and the agent follows next_cursor to the end, so a weekend backlog is judged as one batch. A conversation that already has an assignee or a label is left alone.

  3. 03

    Read each thread

    get_conversation gives the customer's messages and any internal note. The agent decides from the whole thread, not the subject line, and quotes the sentence it relied on.

  4. 04

    Apply your rule

    A thread that fits one topic goes on the pile for that label. A thread that fits two topics, none, or only holds an attachment stays untouched and is listed for you.

  5. 05

    Label, assign, explain

    add_label runs first. It reaches nobody and remove_label undoes it, so it does not wait for your OK. assign_conversation hands the conversation to a colleague and add_internal_note writes the reason, so both ask for your OK by default.

  6. 06

    Check before assigning

    Right before assigning, the agent reads the conversation again. If a colleague took it in the meantime, the agent skips it and logs it as already handled.

What you get back

Example data

New conversations: 14
Labelled and assigned 11:
- 5 x Billing, assigned to the Billing team
- 4 x Delivery, assigned to Sam de Boer
- 2 x Bug, assigned to the Tech team
Left for you 3:
- "Question about a partnership" fits no label
- Two with only an attachment

Other ways to ask

  • Label new conversations about delivery and assign them to the Logistics team, and leave the rest alone.
  • Show me how you would label today's new conversations as a table, but change nothing.
  • Every morning at 8, triage the new conversations and list the ones that fit no label.

What Trengo allows, and what to know first

  • A Trengo label is created for chosen channels under Settings > Settings > Labels, so a label made for one channel may not be offered on another.
  • Trengo lets you archive a label instead of deleting it, so an old label can stay in reports without being added again.
  • A conversation can go to a user or to a team, and a view under Settings > Settings > Views can filter by users, status, labels or channels, so labelled conversations show up where your team looks.
  • A label added to a conversation fires the Ticket label added trigger of Trengo's Zapier app, so an existing Zap can continue from the label.

assign_conversation and add_internal_note ask for your OK by default. To run it daily without you, switch confirmation off for those tools in your dashboard.

Sources: Trengo Help: Adding labels to your inbox · Trengo Help: Assigning a conversation · Trengo Help: Rules explained · Trengo Help: Bulk actions in your inbox · Trengo Help: AI labeling in Journeys · Trengo Help: Adding views to your inbox

Call add_label from a script

curl

curl -X POST https://cotlet.dev/api/v1/trengo/conversations/5120344/labels \
  -H "Authorization: Bearer $COTLET_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"label":"Billing"}'

Example response

{
  "conversation_id": 5120344,
  "labels": [
    "Billing"
  ]
}

MCP: add https://cotlet.dev/mcp/v1 with the same key; the call is trengo_write → add_label. Example data.

Schedule it in n8n, Make, Zapier or cron

Trengo tools this automation uses

Cotlet Trengo tools used to triage new Trengo conversations with AI
Tool Parameters
list_labels none
list_assignees none
list_conversations folder, label, last_message_from, cursor
get_conversation conversation_id
add_label conversation_id, label
assign_conversation conversation_id, user_id, team_id
add_internal_note conversation_id, note
All 13 Trengo tools with descriptions

Frequently asked questions

Is it safe to connect Trengo to Cotlet?

Yes. You sign in to Trengo yourself, in your own Cloud Browser in the EU, and we never save your password. Your agents only get Trengo once you enable it, every background request is listed, and the browser profile is stored encrypted with AES-256-GCM, the cipher 1Password uses for its vaults. The AI model never sees a login, posts and messages wait for your OK, and Disable for agents cuts access in one click. Independent scanners grade cotlet.dev A+ on Qualys SSL Labs, securityheaders.com and Mozilla Observatory; live results are on https://cotlet.dev/security.

Give your agent access to Trengo.

Free 7-day trial with 2,000 credits, no card needed. Cotlet is independent and not affiliated with Trengo.