Skip to main content
SmartlyQ ships a hosted MCP (Model Context Protocol) server. MCP is the open standard that lets AI assistants use external tools. With it, an AI agent can post to your socials, generate content, and read analytics on your SmartlyQ account — just by asking.
Server URL

Prerequisites

  • A SmartlyQ account on a Pro plan or higher (API access requires Pro+).
  • An MCP-capable client: Claude.ai, Claude Desktop, Claude Code, Cursor, ChatGPT, Windsurf, or any MCP client.
The server supports OAuth 2.1 sign-in (MCP spec authorization) — no API key needed. This is the only way to connect from claude.ai custom connectors, and the easiest everywhere else:
  1. Add a custom connector / MCP server with the URL https://mcp.smartlyq.com/mcp.
  2. Your client opens a SmartlyQ sign-in window. Log in, pick the workspace to connect, and tick the permissions you want to grant.
  3. Approve — you’re connected. Revoke anytime from Settings > Developer API (the connection appears as an API key).
Claude Code (CLI)

Connect with an API key

Alternatively, authenticate with a SmartlyQ API key (sqk_live_...) from the Developer Dashboard — with the scopes you need enabled (see Scopes below) — passed as a Bearer token.
Once connected, try a prompt like:
“List my connected social accounts, then draft and schedule a launch post across all of them for tomorrow at 9am.”

Managing connections

Every OAuth connection is listed in the SmartlyQ dashboard under Settings → Developer API → Connections (“Connected AI clients”). Each entry shows the client, the workspace, the permissions you granted, and when it was last used, with a Disconnect button that revokes the app’s access immediately.
Disconnecting is instant and irreversible — the app must sign in again to reconnect. API-key connections are managed separately under the API Keys tab.

Available tools

The full API catalog (operations facade)

Beyond the curated tools above, three facade tools expose every operation in the SmartlyQ API (100+ operations across Chatbot, Media, Short URLs, AI Captain, Account, and more) without flooding the model’s context: The agent discovers what it needs (search_operationsdescribe_operationexecute_operation); scopes, rate limits, and billing apply exactly as for direct API calls.

Tool scopes

Each tool requires the matching scope on your API key. Enable them in the Developer Dashboard. If a key lacks a scope, the tool returns:

Async generation

Content generation tools (generate_article, generate_image, generate_video, text_to_speech) run asynchronously and return a job id. The agent should poll get_job until the job’s status is completed to retrieve the result. See Async jobs. The long SEO lookups (/seo/site-audit, /seo/rank-history, /seo/brand-lookup, /seo/prompt-explorer) are also async on the REST API (202 + job_uid). Over MCP you don’t need to do anything: the MCP server waits for these jobs internally and returns the finished result as a normal tool response (30s-2min). If the wait budget is exceeded you receive the job_uid to poll with get_job.

Webhooks

Subscribe an HTTPS endpoint to events with create_webhook (or POST /webhooks) so your agent reacts to posts publishing, comments arriving, or jobs finishing. Deliveries are HMAC-signed and retried with backoff — see the full Webhooks guide for the event list, signature verification, and retry policy.

Billing

MCP calls are ordinary API calls under the hood, so the same billing and credits and rate limits apply. The MCP layer adds no extra charges and grants no extra access — everything is governed by your API key.

Security

  • Treat your API key like a password. If it leaks, rotate it in the Developer Dashboard.
  • The server forwards your key to the SmartlyQ API per request and stores nothing.