Skip to main content
The Everflow MCP Server is in early access and has two availability gates:
  • MCP is not enabled by default. Email support@everflow.io to request access for your network. Until enabled, all connection attempts will be rejected regardless of your API key.
  • MCP is not available on the European cluster. If your account is hosted on api-eu.eflow.team, the MCP Server is not accessible at this time. Check with your account manager if you’re unsure which cluster your network runs on.
The Everflow MCP Server lets AI agents query your network data through natural language — no code, no API calls, no prompt engineering around raw JSON. Instead of building scripts against the REST API, you connect an MCP-compatible client (Claude, Cursor, VS Code Copilot) directly to Everflow and ask questions in plain English.

​
Server details

The MCP Server is hosted on its own dedicated subdomain, mcp.eflow.team — separate from the main api.eflow.team REST API. All MCP traffic (session init, tool calls, streaming responses) goes through a single streamable HTTP endpoint.
Server URLhttps://mcp.eflow.team
TransportStreamable HTTP
Auth headerX-Eflow-API-Key: <your-network-api-key>
MCP uses your existing Network API key — the same one used for REST API calls, no separate credential. Affiliate and advertiser keys are not accepted. Create or manage keys under Control Center → Security; see Authentication for details. The MCP Server enforces the same permission model as a standard REST request — the authenticated key determines exactly what data the agent can access.

​
What you can do

Once connected, an AI agent can:
  • Pull performance metrics and summaries for any date range
  • Look up offer details, affiliate profiles, clicks, and conversions
  • Diagnose tracking issues by transaction ID
  • Search across your platform data using natural language filters
  • Run multi-step workflows — for example, finding underperforming affiliates, checking their activity, and drafting a summary — without writing a single line of code

​
What it cannot do

The MCP Server is read-only. It cannot create, update, or delete any data. All write operations — creating offers, updating affiliate status, adjusting payouts — must go through the Network API.
The MCP Server is a query interface, not an event system. It does not push data to you, fire webhooks, or trigger on activity in your network. If you ask “tell me when a new affiliate signs up” — that is not something MCP can do. For event-driven workflows, use Webhooks instead.

​
Pagination

List tools and run_performance_report use cursor-based pagination. Every paginated response includes these envelope fields alongside the data array:
FieldTypeDescription
has_morebooleantrue if additional pages exist — pass next_cursor to retrieve them
next_cursorstringOpaque token to pass as cursor in the next request. Absent when has_more is false
rows_returnednumberNumber of records in this page
total_matchingnumberTotal records matching the query across all pages
run_performance_report is an exception: it returns total_rows instead of total_matching. All other paginated tools use total_matching.
search_clicks and search_conversions are not paginated — they return a single capped result set with a capped_at field instead.

​
Limits

Result caps, rate limits, and per-parameter length constraints are documented in Limits & Constraints.

Quickstart

Connect Gemini CLI, Claude Desktop, Cursor, or VS Code in under 5 minutes.

Tools Reference

Every tool, parameter, and response field.

Prompt Library

Ready-to-use prompts for common performance marketing workflows.

Agentic Examples

End-to-end multi-step workflows with full agent traces.

Limits & Constraints

Result caps, rate limits, and parameter length constraints.

FAQ

Common questions about access, data scope, and troubleshooting.

Authentication

How to create and manage Network API keys.