Integration

raia integration

raia hosts AI agents that talk to people over live chat, email, SMS and voice. With one agent's secret key you can prompt the agent, create conversation users and conversations, start an email, SMS or voice conversation with a person, send messages and read the history, take a conversation over manually, manage the agent's knowledge files and vector store, set usage limits, read usage and statistics, and add webhooks. raia allows each organization 1000 requests per 60 seconds for system operations and 300 for AI-backed ones (starting conversations, messages, prompts, knowledge-file and vector-store changes). Not exposed: the organization-key and super-admin routes, the signed upload-URL routes for message and prompt attachments, and a deprecated file list. Authentication is the agent's secret key (in raia CX: open the agent, enable the API Skill, generate a key) sent in the Agent-Secret-Key header, plus the region host (US or EU).

Connect raia How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in raia.Built from the 47 raia tools below, on the connection you add once.

What you can do with raia

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Agent file, 8 operations

add_agent_file_to_vector_storeAdd agent file to vector storePUTAdds an uploaded knowledge file to the agent's vector store and waits for it, so the agent can search it.
add_agent_file_to_vector_store_asyncAdd agent file to vector store asyncPUTStarts adding an uploaded knowledge file to the agent's vector store and returns before it is done; check vectorStoreUploadStatus with get_agent_file.
delete_agent_fileDelete agent fileDELETEDeletes one of the agent's knowledge files by id. Irreversible.
delete_agent_filesDelete agent filesDELETEDeletes several of the agent's knowledge files by id. Irreversible.
get_agent_fileGet agent fileGETReturns one of the agent's knowledge files by id: name, URL, created date and vector-store status (processing, done or failed).
list_agent_filesList agent filesGETLists the agent's knowledge files (name, URL, created date, vector-store status), paged from page 0, sorted by createdAt or fileName.
remove_agent_file_from_vector_storeRemove agent file from vector storeDELETERemoves a knowledge file from the agent's vector store, so the agent no longer searches it; the file itself stays.
upload_agent_fileUpload agent filePOSTUploads a knowledge file to the agent and, unless autoAddToVectorStore is false, adds it to the agent's vector store so the agent can search it. Returns the file with its id and vector-store status.

Conversation, 5 operations

create_conversationCreate conversationPOSTCreates a conversation between the agent and an existing conversation user, without contacting the person. Returns the conversation with its id; send messages with send_message_to_agent.
delete_user_conversationsDelete user conversationsDELETEDeletes ALL conversations of one conversation user with this agent. Irreversible.
get_conversationGet conversationGETReturns a conversation by id: title, summary, score, analysis status, mode, state, user and assigned team member (personal data), and its feedbacks.
list_conversationsList conversationsGETLists this agent's conversations with title, summary, score, mode, state and the user (personal data). Filter by keyword, user, updated-date range or feedback; paged.
update_conversationUpdate conversationPUTChanges a conversation's title, context, status (active, inactive, archived) or state; fields left out keep their value. Returns the conversation.

Agent limits, 4 operations

create_agent_limitsCreate agent limitsPOSTSets the agent's monthly limits: tokens, SMS, emails and voice seconds per month (all four required).
delete_agent_limitsDelete agent limitsDELETERemoves the agent's own monthly limits.
get_agent_limitsGet agent limitsGETReturns the agent's monthly limits: tokens, SMS, emails and voice seconds, and whether they are the organization default.
update_agent_limitsUpdate agent limitsPUTChanges the agent's monthly limits; each of tokens, SMS, emails and voice seconds left out keeps its current value.

User, 4 operations

create_userCreate userPOSTCreates a conversation user (a person the agent talks to) from a first and last name plus at least one of email, phoneNumber or fkId (your own id for the person). Returns the user with its id.
delete_userDelete userDELETEDeletes a conversation user by id. Irreversible.
find_userFind userGETFinds a conversation user by email, phone number or fkId and returns the user (personal data) with the ids of their conversations with this agent. raia answers 404 when no user matches.
get_userGet userGETReturns a conversation user by id: name, email, phone number, fkId, context, source and metadata (personal data).

User limits, 4 operations

create_user_limitsCreate user limitsPOSTSets monthly limits for one conversation user: tokens, SMS, emails and voice seconds per month (all four required).
delete_user_limitsDelete user limitsDELETERemoves a conversation user's own monthly limits.
get_user_limitsGet user limitsGETReturns a conversation user's monthly limits: tokens, SMS, emails and voice seconds, and whether they are the organization default.
update_user_limitsUpdate user limitsPUTReplaces a conversation user's monthly limits: all four values are required, so read them first with get_user_limits and send back the ones you keep.

Agent message, 2 operations

send_message_to_agentSend message to agentPOSTSends a message to the agent in a conversation, as the conversation user, and waits for the agent's answer, returned with its citations and token usage. Uses the agent's tokens.
send_message_to_agent_asyncSend message to agent asyncPOSTSends a message to the agent in a conversation and returns the stored user message at once (its id). Call get_message_response with that id until status is completed or failed.

Prompt, 2 operations

create_prompt_jobCreate prompt jobPOSTSends one prompt to the agent outside any conversation and returns promptId at once. Call get_prompt_response with it until the answer is there.
run_promptRun promptPOSTSends one prompt to the agent outside any conversation and waits for the answer: status, response text, generated artifacts and token usage. Uses the agent's tokens.

Agent, 1 operation

get_agentGet agentGETReturns the agent this key belongs to: name, public name, role, description, status, tags, organization, vector store id and AI settings (model, temperature, instructions). Skill settings are not returned.

Agent admin, 1 operation

list_agent_adminsList agent adminsGETLists the platform users who administer this agent (direct owners or admins and organization owners or admins): email, name, roles, status and access source (personal data), paged.

Agent statistics, 1 operation

get_agent_statisticsGet agent statisticsGETReturns the agent's usage statistics over a date range, optionally for one conversation user: tokens used and per conversation, voice seconds and average call length, SMS and emails sent and per conversation.

Agent usage, 1 operation

get_agent_usageGet agent usageGETReturns the agent's resource usage: emails, SMS, tokens and voice seconds.

Agent user, 1 operation

list_agent_usersList agent usersGETLists the conversation users of this agent (name, email, phone number, fkId: personal data), paged.

Agent webhook, 1 operation

create_agent_webhookCreate agent webhookPOSTAdds a webhook to the agent: raia POSTs to url on triggerCondition (each message, start or end of dialog and others). The url must be public HTTPS. raia documents no delivery signature: authenticate with headers.

Conversation message, 1 operation

list_conversation_messagesList conversation messagesGETReturns every message of a conversation in order (oldest first by default): text, sender role, source channel, time, the AI's full answer with citations, token usage, functions used, and optionally feedbacks.

Conversation mode, 1 operation

update_conversation_modeUpdate conversation modePUTSwitches a conversation between ai (the agent answers) and manual (a human answers with send_manual_message; the agent stays silent). Returns the conversation.

Conversation state, 1 operation

list_conversation_statesList conversation statesGETLists the conversation states defined for this agent (id, name, description, color, persistent flag), to set on conversations through conversationStateId.

Email conversation, 1 operation

start_email_conversationStart email conversationPOSTStarts a conversation with a person by EMAILING them now: the agent sends emailIntroduction, under emailSubject, to email. Sends a real email. Returns conversationId.

Manual message, 1 operation

send_manual_messageSend manual messagePOSTSends a message from a human team member to the person in a conversation in manual mode, over the conversation's channel (not voice). Delivers a real message to the person.

Message feedback, 1 operation

save_message_feedbackSave message feedbackPOSTCreates or replaces a user's up or down vote (with question, answer and comment) on one agent message. Returns the feedback.

Message file, 1 operation

get_message_file_download_urlGet message file download urlGETReturns a signed URL to download one file attached to a conversation message (the file must belong to this agent's conversation).

Message response, 1 operation

get_message_responseGet message responseGETReturns the agent's answer to one user message: status pending, completed or failed, the answer message when completed, and error when failed. Call again while pending.

Prompt response, 1 operation

get_prompt_responseGet prompt responseGETReturns the answer to a prompt from create_prompt_job: status, response text, artifacts and token usage. raia returns nothing (an empty result) while the prompt is not ready: call again.

Resource usage, 1 operation

log_resource_usageLog resource usagePOSTRecords resource usage on the agent (LLM tokens or voice seconds you consumed outside raia), with optional model details. It counts toward the agent's usage and limits.

Sms conversation, 1 operation

start_sms_conversationStart sms conversationPOSTStarts a conversation with a person by TEXTING them now: the agent sends smsIntroduction by SMS to phoneNumber. Sends a real SMS. Returns conversationId.

Voice conversation, 1 operation

start_voice_conversationStart voice conversationPOSTStarts a conversation with a person by CALLING them now: the agent phones phoneNumber and opens with firstMessage. Places a real phone call. Returns conversationId.

Use raia from Claude Code, Cursor or Codex

Connect raia in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its raia operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect raia

Create your workspace
Sign up free. Nothing to install.
Connect raia once
Add your raia connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.