Model Context Protocol (MCP)
Connect Outercite to MCP-compatible AI assistants so they can query your citation data, competitors, and agency portfolio directly.
The Outercite MCP server lets MCP-compatible AI assistants (such as Claude) read your Outercite data directly, so you can ask questions like "what is my citation rate this week?" or "which competitors are gaining on Perplexity?" right inside your assistant. It is the natural companion to a product about AI search: the same engines that cite your brand can now query how you are being cited.
What you'll learn
- What the Model Context Protocol is and why Outercite exposes one
- Which tools the Outercite MCP server provides
- How to connect it to your MCP client
What is MCP?
The Model Context Protocol is an open standard for connecting AI assistants to external tools and data. An MCP server exposes a set of typed tools; an MCP client (your AI assistant) can call them on your behalf. Outercite ships an MCP server so your assistant can pull live citation analytics without you opening the dashboard.
MCP access uses the same API key as the Outercite API. Generate or manage keys under Settings. Treat the key like a password: anyone with it can read your workspace data.
Tools the server provides
The server groups its tools by area. Read-only tools fetch data; a few write tools let an assistant take action when you ask it to.
Workspace tools (outercite_*)
| Tool | What it does |
|---|---|
get_dashboard_stats | Rolled-up citation rate, visibility score, 7-day delta, pending actions |
get_platform_coverage | Per-engine citation rate and trend across the six engines |
list_keywords | Tracked prompts with visibility, position, and platform coverage |
list_competitors | Competitors with share of voice, threat score, and threat level |
list_citations / get_citation | Verified citations with tier, confidence score, and proof snippet |
list_insights | Generated insights and their attached actions |
list_actions / mark_action_complete | Action items and their status |
get_brief_latest | The latest generated brief for the workspace |
add_keyword / add_competitor | Add a tracked prompt or competitor |
trigger_audit | Kick off a fresh verification run |
create_webhook / list_webhooks / delete_webhook | Manage webhooks |
Agency tools (agency_*)
| Tool | What it does |
|---|---|
list_clients | All client workspaces with visibility and activity |
portfolio_health | Clients bucketed as Needs Attention, Monitoring, or On Track |
compare_clients | Side-by-side comparison across clients |
generate_client_brief | A brief for a single client |
recent_activity | Recent citation activity across the portfolio |
Agency tools are available to agency (parent org) accounts. Scope a request to one client by passing its workspace id, or omit it to work across the whole portfolio.
Connecting the server
You connect the Outercite MCP server in your AI assistant's MCP settings. The exact steps depend on the client, but the shape is the same:
- Get an API key from your Outercite account under Settings.
- Add the Outercite MCP server to your client's MCP configuration and supply the key.
- Reload your client. The
outercite_*(and, for agencies,agency_*) tools become available.
{
"mcpServers": {
"outercite": {
"url": "https://mcp.outercite.ai",
"headers": { "Authorization": "Bearer <your-api-key>" }
}
}
}The configuration above is an illustrative example. Use the exact server address and connection method shown in your Outercite account, which may change as the MCP server evolves.
Try this in Outercite
Generate an API key in your account settings, add the Outercite server to your assistant, then ask it: "What is my citation rate across all six engines this week?"