Crisp connector for AI agents

Tools your support agent reaches for on Crisp, scoped per user.

CALL ANY TOOL

50 tools for live chat in Crisp: list and read conversations and messages, send replies, assign, route, and resolve chats, manage contact profiles, data, and events, read helpdesk pages, and generate analytics.

crisp_add_people_event

Add people event

Record an event on a contact's timeline, for example a purchase or signup. Returns an empty data object on success. Use crisp_list_people_events to read the timeline back.

Parameters

Name

Type

Required

Description

people_id

string

Required

People profile identifier from crisp_list_people_profiles or crisp_get_conversation (people_id). Crisp also accepts the contact's email address in place of the identifier. Example: 3b6f2a10-7c9d-4e1a-8f55-0a1b2c3d4e5f.

text

string

Required

Short description of the event. Example: Upgraded to Pro plan.

website_id

string

Required

Identifier of the Crisp website (workspace) to act on. Find it in the Crisp app under Settings > Workspace Settings > Setup Instructions, or call crisp_get_website on a known id. Example: 8c042840-1ecb-4cd5-a6a1-b3ac4a3ad8d6.

color

string

Optional

Color label for the event in the Crisp UI. One of: `red`, `orange`, `yellow`, `green`, `blue`, `purple`, `pink`, `brown`, `grey`, `black`.

data

object

Optional

Extra key-value data for the event. Example: {"plan": "pro"}.

crisp_assign_conversation_routing

Assign conversation routing

crisp_create_conversation

Create conversation

crisp_delete_conversation

Delete conversation

crisp_generate_analytics

Generate analytics

crisp_get_availability_status

Get availability status

crisp_get_conversation

Get conversation

crisp_get_conversation_participants

Get conversation participants

crisp_get_conversation_routing

Get conversation routing

crisp_get_helpdesk

Get helpdesk

crisp_get_inbox

Get inbox

crisp_get_message

Get message

crisp_get_people_data

Get people data

crisp_get_people_stats

Get people stats

crisp_get_people_subscription

Get people subscription

crisp_get_website

Get website

crisp_list_conversation_events

List conversation events

crisp_list_conversation_files

List conversation files

crisp_list_helpdesk_locales

List helpdesk locales

crisp_list_inboxes

List inboxes

crisp_list_operator_availabilities

List operator availabilities

crisp_list_operators

List operators

crisp_list_people_events

List people events

crisp_list_teams

List teams

crisp_mark_messages_read

Read mark messages

crisp_send_message

Send message

Send a message into an existing conversation as the operator or on behalf of the visitor. Returns the message fingerprint inside the data object once the message is accepted. Use type note with from operator for a private internal note. Use crisp_create_conversation first when no conversation exists.

Parameters

Name

Type

Required

Description

content

string

Required

Message body. For type text or note this is a plain string, for example Thanks for reaching out. For file, animation or audio it is an object with url, type and name (plus duration for audio). For picker, field, carousel and event it is an object as described in the Crisp REST API reference.

from

string

Required

Who the message comes from: operator for a reply from the team, or user for the visitor. One of: `user`, `operator`.

origin

string

Required

Channel the message belongs to: chat for the Crisp chatbox, email for email, or a urn: value for a custom integration. Match the conversation's origin.

session_id

string

Required

Session identifier of the conversation, as returned in session_id by crisp_list_conversations or crisp_create_conversation. Example: session_6c1d8b2a-3f4e-4a5b-9c7d-1e2f3a4b5c6d.

type

string

Required

Kind of message. Use text for a normal message, note for a private operator note, or file, animation, audio, picker, field, carousel or event for rich content. One of: `text`, `note`, `file`, `animation`, `audio`, `picker`, `field`, `carousel`, `event`.

website_id

string

Required

automated

boolean

Optional

Mark the message as automated, meaning it comes from a bot.

fingerprint

number

Optional

Client-chosen unique numeric fingerprint, useful to deduplicate messages. Crisp generates one when omitted.

mentions

array

Optional

User or team identifiers to mention in the message. Mainly used with notes.

original

object

Optional

Original message data to keep, with type (text/plain or text/html) and content. Example: {"type": "text/plain", "content": "Hello"}.

properties

object

Optional

Arbitrary key-value properties stored on the message.

references

array

Optional

Extra context for the message. Each item has type (link), name and target URL. Example: [{"type": "link", "name": "Docs", "target": "https://example.com/docs"}].

stealth

boolean

Optional

When true the message is stored but not propagated to the other party.

timestamp

number

Optional

Send time in milliseconds when it should differ from now, for example to import older messages.

translated

boolean

Optional

Mark the message as auto-translated.

user

object

Optional

Display identity for the sender, with optional keys type (website or participant), nickname and avatar. Example: {"nickname": "Support Bot"}.

crisp_update_conversation_inbox

Update conversation inbox

crisp_update_conversation_meta

Update conversation meta

crisp_update_message

Update message

crisp_update_people_profile

Update people profile

For more tools, view docs.

Build your Agent

Same auth pattern across LangChain, OpenAI, Anthropic, and Google ADK.

import { ScalekitClient } from "@scalekit-sdk/node";
import { createReactAgent } from "@langchain/langgraph/prebuilt";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

// Crisp tools scoped to this user
const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["crisp"], toolNames: [
    "crisp_list_conversations",
    "crisp_list_messages",
    "crisp_send_message"] },
  pageSize: 100,
});

const agent = createReactAgent({ llm, tools });
await agent.invoke({ messages: [{ role: "user", content: "Reply to unresolved chats that ask about refund timing" }] });

import OpenAI from "openai";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const openai = new OpenAI();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["crisp"] }, pageSize: 100,
});

const res = await openai.chat.completions.create({
  model: "gpt-5",
  messages: [{ role: "user", content: "Reply to unresolved chats that ask about refund timing" }],
  tools,
});

// Execute the tool call with the user's vaulted Crisp website token (basic auth)
await sk.tools.executeTool(res.choices[0].message.tool_calls[0], "user_123");

import Anthropic from "@anthropic-ai/sdk";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const anthropic = new Anthropic();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["crisp"] }, pageSize: 100,
});

const msg = await anthropic.messages.create({
  model: "claude-sonnet-5",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Reply to unresolved chats that ask about refund timing" }],
  tools,
});

// Tool call runs with the user's vaulted Crisp website token (basic auth)
await sk.tools.executeTool(msg.content, "user_123");

import { Agent } from "@google/adk/agents";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["crisp"] }, pageSize: 100,
});

const agent = new Agent({
  name: "support_agent",
  model: "gemini-2.5-pro",
  instruction: "Crisp tools scoped to this user",
  tools,
});

await agent.run("Reply to unresolved chats that ask about refund timing");

Try these prompts

Copy any prompt into your agent. Each maps directly to a Crisp tool. Click to copy, paste into your agent, done.

Conversations

Copy the prompt

Copied

List unresolved conversations from the last 24 hours.

Copy the prompt

Copied

Reply to the latest chat from jane@acme.com with our refund timeline.

Copy the prompt

Copied

Mark the Globex conversation as resolved.

Routing and operators

Copy the prompt

Copied

Assign the oldest unassigned chat to Maria.

Copy the prompt

Copied

Show which operators are available right now.

Copy the prompt

Copied

Move this conversation to the Billing inbox.

Contacts and insights

Copy the prompt

Copied

Look up the contact profile for alex@globex.com.

Copy the prompt

Copied

Record a Purchased Pro event on Priya Shah's timeline.

Copy the prompt

Copied

How many visitors are on the website right now?

SEE HOW AUTH WORKS

Each user connects their own Crisp website token once, and Scalekit sends it with every call. Credentials stay vaulted, every call is scope checked, and every action is logged.

Authorize

Your user connects

Crisp

once. We tie it to their identity and the meetings they approved — no shared bot account, no org-wide access

Who:

user ‘A’

when:

Once per user

access:

Limited to user

Store

Their

Crisp

token lives in a vault scoped to them. User A's meetings are never reachable by an agent acting for user B, even on the same connection

vault:

encrypted

scope:

per-user

tokens:

auto-refreshed

Resolve

When your agent calls a

Crisp

tool, we fetch the right token server-side. It never touches your agent, never appears in the LLM context, never shows up in your logs

speed:

~40ms

check:

before every call

seen by:

nobody

Audit

Every

Crisp

tool call is logged — who triggered it, which meeting was fetched, what came back. 90 days of history, tied to the user who authorized it

history:

90 days

export:

SIEM-ready

logged:

every call

Test other agents

See the same per-user auth pattern across other customer support connectors.

Support and Ops Teams

Support triage agent

Fetches new Zendesk tickets, classifies them by type and urgency, searches the Notion knowledge base for an answer, and routes what it cannot resolve to Slack. Every call runs on the support agent's own delegated OAuth.

Support and Ops Teams

Support ticket automation agent

Fetches new Zendesk tickets, drafts a reply from Notion knowledge base articles, digests what it cannot answer to Slack, and archives the rest, acting as the support agent rather than a shared API key.

Why Scalekit

Secure your agent's access. Connectors ship in minutes

01.

Shared tokens break per-user analytics

A shared Crisp website token looks fine in a demo. In production every reply and status change looks like one service account, and you cannot tell which user triggered it. Scalekit resolves the credential of the actual user who triggered the agent, never a shared bot.

// shared token
audit → bot_service_account

// scalekit
audit → user_abc ✓

02.

Authentication is not authorization

03.

Multi-tenancy is architectural

04.

Crisp today. Ten connectors tomorrow.

“Our agents act across Salesforce, Gong, Google Drive, and more, on behalf of every customer. Scalekit behind the scenes meant we can keep adding tools without ever rebuilding how credentials or tool calling work.”

Venu Madhav Kattagoni

Head of Engineering / Von

FAQs

Frequently Asked Questions

Does the agent access Crisp as the user or through a shared website token?

As the user. Each user connects their own Crisp website token once, and Scalekit sends it with every call. Audit logs attribute every reply and conversation change to that user, not a shared service account.

Where are the Crisp credentials stored?

In Scalekit's managed AES-256 token vault, namespaced per tenant. Revocation is a single dashboard action. Credentials never appear in prompts, logs, or LLM context.

Can I limit what the agent does in Crisp?

Yes. Filter by tool name in listScopedTools to expose only what you want, for example the conversation and contact read tools without crisp_send_message or crisp_delete_conversation, which permanently removes a conversation and its messages. Scalekit also enforces scope checks before every API call.

What happens when Crisp access is turned off?

Disable the connection in the Scalekit dashboard and the next tool call fails closed with a clear error. Other tenants are unaffected, and the change is logged.

Can the agent use our Crisp helpdesk articles?

Yes. crisp_list_helpdesk_pages lists published helpdesk pages for a locale, and crisp_get_helpdesk_page_entity resolves one page, so an agent can find the right article before it replies in a conversation.