Skip to content
Agent-native customer support

Onboard your customers with zero UI

Your agent connects to a Corebee org, ingests the customer docs, brands the widget, and runs support end to end through one MCP and REST API. No dashboard. No clicks.

Set up from your terminalTalk to us about agents
  • MCP-native
  • OAuth 2.1 + scoped keys
  • Free plan, no card
connect.sh

# resolve the org this key is scoped to

GET /api/v1/me

200 OK

{
  "organization": {
    "id": "org_a1c8...",
    "name": "Acme Inc"
  }
}
Quickstart

Run a customer's support in a handful of calls

The same workspace, whether your agent speaks plain REST or the Model Context Protocol. Real endpoints, real tool names, no inferred call order.

rest-v1.sh
# 1. Verify the key and resolve the org it is scoped tocurl https://corebee.ai/api/v1/me \  -H "Authorization: Bearer $COREBEE_API_KEY"# 200 -> { "organization": { "id": "...", "name": "Acme Inc" } }# 2. Create a contact for the customer's end-usercurl -X POST https://corebee.ai/api/v1/contacts \  -H "Authorization: Bearer $COREBEE_API_KEY" \  -H "Content-Type: application/json" \  -d '{ "email": "buyer@acme.com", "name": "Dana Buyer" }'# 3. Read the support conversations the AI is handling#    (status is an integer code; see the OpenAPI spec for the enum)curl "https://corebee.ai/api/v1/conversations?limit=20" \  -H "Authorization: Bearer $COREBEE_API_KEY"
mcp-client.ts
// Same workspace, over the Model Context Protocol.// Connect: https://corebee.ai/api/v1/mcp/stream  (OAuth 2.1, scoped per org)// Ingest the customer's docs into the knowledge baseawait mcp.callTool("add_knowledge_url", {  url: "https://acme.com/docs",});// Brand the widget for the customerawait mcp.callTool("configure_widget_appearance", {  widget_color: "#2E7D6B",  position: "right",});// Read how the AI support is performingconst kpis = await mcp.callTool("get_metrics", {});
Discovery

The surfaces your agent can read

Every artifact is static, discoverable, and points to the next. Discovery leads to the contract, the contract leads to the tools.

llms.txt

https://corebee.ai/llms.txt

Discovery breadcrumb. Brand facts plus pointers to the API base URL, OpenAPI spec, agent manifest, and MCP endpoint.

Agent manifest

https://corebee.ai/.well-known/agents.json

Links plus the connect, ingest-KB, configure-widget, operate flow, so the agent does not have to infer the call order.

MCP server card

https://corebee.ai/.well-known/mcp/server-card.json

Points at the live Corebee MCP endpoint and its OAuth protected-resource metadata. Tools are typed and self-describing.

OAuth metadata

https://corebee.ai/.well-known/oauth-protected-resource

RFC 9728 discovery. Leads to the authorization server, Dynamic Client Registration, and per-org scoped tokens.

OpenAPI spec

https://corebee.ai/api/v1/openapi.json

The v1 REST contract: typed request and response schemas for conversations, contacts, labels, and webhooks.

MCP endpoint

https://corebee.ai/api/v1/mcp/stream

The live Model Context Protocol server. Connect any MCP-capable agent and call typed Corebee tools directly. The server card lists the current tools.

The sequence

How an agent onboards a customer

Five steps, no human in the loop. Discover, authenticate, connect, ingest, deploy.

  1. 01

    Discover

    The agent fetches llms.txt, the agent manifest, and the MCP server card to learn the endpoints, flows, and auth requirements. No human reads docs.

  2. 02

    Authenticate

    OAuth 2.1 authorization code with PKCE, self-registered through Dynamic Client Registration, or a scoped API key. The agent gets a token bound to a single org.

  3. 03

    Connect to the org

    Verify the credential and resolve the workspace with GET /api/v1/me. Every call from here is scoped to that one customer org. Nothing leaks across tenants.

  4. 04

    Ingest the knowledge base

    Point Corebee at the customer docs URL with add_knowledge_url. RAG indexing builds the answer base, which resolves about 86% of questions automatically.

  5. 05

    Configure and operate

    Brand the widget with configure_widget_appearance, then read conversations and metrics and reply, all through the API. A live AI support widget is answering, no dashboard touched.

Live today

Everything above runs against a connected Corebee org with no UI: ingest the knowledge base, brand the widget, read conversations and metrics, and reply, all through the REST v1 API and the MCP server. The org is created once through signup or our partner flow.

Early access

Fully autonomous account provisioning, where your agent creates a brand-new customer org and subscribes it to a paid plan without a human, is in early access. If you are building that, tell us about your use case and we will get you set up.

Why build on Corebee

Designed for the agent that installs it

An API-first help desk built for the agent doing the work, not the human who never logs in.

MCP-native

A live Model Context Protocol server, discoverable through well-known metadata. Connect any MCP-capable agent and call typed tools directly.

One flow, not a pile of endpoints

The agent manifest bundles the connect, ingest, and deploy sequence with flows and links so your agent succeeds on the first try.

Operate a real workspace, not a sandbox

Calls hit the same live org your customer uses. Ingest docs, brand the widget, and reply for real, with a widget answering customers in minutes.

Flat pricing, nothing to reason about

Free to start. Pro is $99/mo with unlimited resolutions, seats, and conversations. No per-seat math for your agent to model.

86% auto-resolution out of the box

RAG retrieval over the ingested docs resolves the bulk of questions with no tuning. Your agent ships support that actually works.

Audit logs that separate agent and human

Every action is attributed. You can tell exactly what your agent did versus what a person did, with a full trail per org.

Bounded autonomy

Security and scoped permissions

Letting an agent run support is only safe when the blast radius is bounded. Corebee scopes it on every axis.

Two human-absent auth paths

  • OAuth 2.1 authorization code with PKCE, self-registered via Dynamic Client Registration. Discovered through RFC 9728 then RFC 8414.
  • Scoped API keys for agents that cannot speak OAuth, bound to a single org.
  • Short-lived per-org tokens. Revoke at any time, instantly.

Bounded autonomy

  • Per-action scopes. A read token can never write or provision.
  • Human confirmation required for destructive operations by default.
  • Rate-limited writes that fail closed, so a runaway agent cannot flood an org.
  • SOC 2 in progress, GDPR compliant, data encrypted at rest and in transit.
Tool catalog

A representative slice of the typed tools. Each tool declares its access level so your agent can reason about scope before it calls.

add_knowledge_urlwriteCrawl and index a docs URL into the RAG knowledge base.
search_knowledgereadQuery the indexed knowledge base and return matching chunks.
configure_widget_appearancewriteSet widget color, logo, position, and welcome text.
list_conversationsreadList support conversations and their AI resolution status.
reply_to_conversationwriteSend a reply to a conversation, visible to the customer.
get_metricsreadConversation, message, contact and knowledge-base counts for a period, plus CSAT and first-response time.
create_contactwriteCreate an end-user contact record.
delete_knowledge_sourcedestructiveRemove a source and its indexed content. Requires confirmation.
The agent economy

The first help desk an agent can install

End to end, for its own customer. No human in the loop.

MCP
Typed tools over one endpoint, listed live in the server card
$0/mo
Flat pricing, no per-seat math to reason about
0%
Average auto-resolution out of the box

FAQ

Got Questions? We've Got Answers.

What can an agent do through the Corebee API today?
Against a connected Corebee org, an agent can ingest a knowledge base, brand the widget, read conversations and metrics, and reply, all through the REST v1 API and the MCP server with no dashboard. The org is created once through signup or our partner flow.
How does an agent authenticate to Corebee?
Two human-absent paths. OAuth 2.1 authorization code with PKCE, self-registered through Dynamic Client Registration and discovered via RFC 9728 then RFC 8414, or a scoped API key bound to a single org. Tokens are short-lived and revocable, and every call is scoped to one customer org.
Does Corebee have a Model Context Protocol server?
Yes. The live MCP endpoint is https://corebee.ai/api/v1/mcp/stream over streamable HTTP. Connect any MCP-capable agent and call typed Corebee tools directly. The current tool list is in the server card at https://corebee.ai/.well-known/mcp/server-card.json.
How much does Corebee cost for an agent to deploy?
Three flat plans. Free is $0 with 50 AI conversations a month on web chat. Pro is $99 per month with unlimited AI conversations (fair use) and unlimited seats. Business is $199 per month. There is no per-seat math and no per-resolution fee for your agent to model. Pro and Business include a 14-day free trial (card required) and a 30-day money-back guarantee.
Can an agent create a brand-new customer account on its own?
Yes — live today, with exactly one human checkpoint. The agent calls POST /api/auth/agent-signup with the user’s email; Corebee emails a confirm link carrying a pairing phrase the agent also shows in the terminal. The user clicks Approve once (any device), and the agent polls the returned poll_url to receive a scoped 24-hour API key, the provisioned org, the widget embed snippet, and an already-running docs crawl. No dashboard visit is required to reach a working support widget. For security, emails that already have a Corebee account get no auto-issued key — the signed-in user creates one from the dashboard. Paid plan subscription also stays in the dashboard.
How is auto-resolution achieved without tuning?
RAG retrieval over the ingested docs resolves about 86% of questions automatically out of the box. Point Corebee at the customer docs URL with add_knowledge_url and the answer base builds itself.

Ship support your agent installs itself

Read the docs, wire up the flow, and let your agent onboard customers end to end. We are happy to help you design the integration.

Read the API docsTalk to us about building on Corebee
  • MCP-native
  • OAuth 2.1 + scoped keys
  • Free plan, no card