MCP Server

The Genudo Console exposes a Model Context Protocol (MCP) server. MCP is an open standard that lets AI clients (Claude, ChatGPT, Codex, Gemini CLI, Cursor, Windsurf, and any MCP-compatible agent) discover and invoke tools at runtime over a persistent connection.

Your token can invoke tools that manage pipelines, stages, actions, variables, opportunities, contacts, messages, and analytics — all as the authenticated user, respecting the same permission rules as the REST API.

Endpoints

Endpoint	Purpose
POST /mcp	JSON-RPC 2.0 — all MCP requests (initialize, tools/list, tools/call)
DELETE /mcp	Terminate a session

The transport is Streamable HTTP (MCP spec 2025-03-26): every request is a stateless POST with a JSON-RPC 2.0 body, every response is application/json. The MCP handshake, tool discovery, and per-tool invocations all flow through your client library — you rarely call these endpoints by hand. See Connect.

Authentication

Two ways in, both landing on the same mcp:use scope:

• OAuth (default) — your client hits the endpoint, gets a 401 with discovery headers, registers itself, and opens a browser sign-in. Approve access and you're connected. No token to create or paste. This is what every path on Connect a client uses.
• Bearer token — send Authorization: Bearer <token> for custom agents and SDKs that don't do OAuth discovery. Tokens are created from the Authentication guide and can be scoped, revoked, and expired independently per integration.

Tools inherit whatever scopes your token carries. A token with mcp:use + opportunities:read can call tools that read opportunities but not those that write. A token with only mcp:use can call tools that require no additional scope.

See Authentication for token creation and scope definitions.

Available tools

32 tools across 5 groups.

Analytics

Account health, messaging volume, AI performance & cost.

Tool	Description
get_account_summary	Account snapshot: total pipelines, opportunities, and message counts
get_messaging_stats	Messaging volume filtered by date range, pipeline, or channel provider
get_ai_performance	Total AI messages, average response time, and total cost

Discover

Pull and filter your workspace data — the starting point for everything.

Tool	Description
list_pipelines	Your pipelines with status, description, and stage counts
list_pipeline_stages	Stages of a pipeline in stage order
list_actions	Actions for a pipeline, optionally filtered to a specific stage
list_action_results	Execution log for a pipeline's actions — did the webhook fire, and what came back
list_variables	Pipeline variables available for use in actions
list_contacts	Contacts with filters for status, priority, tags, and search
list_opportunities	Opportunities with contact, stage, and tags context
list_messages	Messages across pipelines, cursor-paginated
list_knowledge_tables	Knowledge tables with column counts and the pipelines they are attached to
search_knowledge_table	Search a knowledge table by natural-language query
list_knowledge_points	Rows in a knowledge table with their default_ids, paginated
list_model_pool	Current Smart Routing tier assignments for a pipeline
get_stage_followup	Followup configuration and interval schedule for a stage
get_pipeline_options	Valid agent types, AI providers, models, languages, dialects, and channels

Build

Create pipelines, stages, variables, webhook actions, followups, and knowledge tables.

Tool	Description
start_pipeline_journey	Guided builder — the intended first call for a new pipeline
create_pipeline	New pipeline with agent type, AI model, language, persona, and instructions
create_stage	Add a neutral, won, or lost stage with its own AI behavior
create_variable	Define fixed, system, action, or AI-extracted variables
create_action	Outbound HTTP webhook action for a pipeline or stage
create_followup	Re-engagement sequence for a stage with timed intervals
create_knowledge_table	Structured knowledge base table with columns
upsert_knowledge_points	Insert or update rows in a knowledge table

Update

Tune live pipelines, stages, actions, deals, followups, and Smart Routing.

Tool	Description
update_pipeline	Change settings, persona, AI model, limits, and which knowledge tables the agent can search
update_model_pool	Set the four Smart Routing tiers on an existing pipeline
update_stage	Edit instructions, entry conditions, order, or opening messages
update_action	Tune an existing action's request or trigger
update_variable	Edit a variable's value or metadata
update_opportunities	Bulk update status, stage, tags, or notes
update_followup	Change followup intervals, instructions, or activation status

Delete

Tool	Description
delete_knowledge_points	Remove specific rows from a knowledge table by their default_id values

For full parameter reference see Tools. The plugin adds skills and agents on top of these tools — see Skills & agents.

When to use MCP vs the REST API

• REST: one-shot integrations, webhooks, cron jobs, server-to-server automation.
• MCP: interactive AI clients that discover capabilities at runtime and stream results back to a chat interface.

You can use both from the same PAT.