Crisp for AI agents
Crisp is a customer messaging platform for support conversations, letting a team manage chats, contacts, and helpdesk content from one inbox. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Crisp.
Reply to a conversation
Send a message in an active conversation to answer a customer's question.
Look up a customer's profile
Look up a person's profile before responding, to see their past conversations.
Assign a conversation to an operator
Assign a conversation to the right operator so it doesn't sit unanswered.
12 Crisp actions, graded by risk.
Every Crisp action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
7Look things up. Allowed by default.
- crisp.list_inboxesList Inboxes
- crisp.list_operatorsList Operators
- crisp.get_conversationGet Conversation
- crisp.get_people_profileGet People Profile
- crisp.list_conversationsList Conversations
- crisp.list_people_profilesList People Profiles
- crisp.list_conversation_messagesList Conversation Messages
write
5Create and change things. Allow, or ask the user first.
- crisp.update_conversation_metaUpdate Conversation Meta
- crisp.send_conversation_messageSend Conversation Message
- crisp.update_conversation_stateUpdate Conversation State
- crisp.assign_conversationAssign Conversation
- crisp.move_conversation_to_inboxMove Conversation to Inbox
destructive
0Delete, cancel or archive. Ask first, or deny outright.
- No destructive actions.
Crisp in three steps.
- 01Your users connect CrispThey add their Crisp api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, and writes like “send Conversation Message” can wait for the user to approve.
- 03Any agent can actYour agent calls Crisp through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('crisp', { read: 'allow', write: 'ask', // send_conversation_message destructive: 'deny', }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Crisp connects.
Users add their Crisp api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same Crisp connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Crisp from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
Crisp and Arc0, answered.
Can I use Crisp with Claude, ChatGPT or Cursor?
Yes. Connect Crisp to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Crisp actions you allow.
How do users connect Crisp?
Users add their Crisp api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which Crisp actions can my agent take?
12 in total: 7 read, 5 write and 0 destructive, such as “send Conversation Message”. Your policies decide which of them each agent may call.
Can I make my agent read-only in Crisp?
Yes. Allow read actions and deny writes in the Crisp policy. Your agent can still look things up, and any write it attempts is blocked and logged.
Can my own backend call Crisp too?
Yes. The same Crisp connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Crisp into your agent.
Your users connect Crisp once, under your brand. Your agent gets 12 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan