Anthropod Documentation

Connect your AI assistant to Anthropod through Model Context Protocol (MCP). Ask questions about conversation analytics, customer journeys, agent performance, and Voice Agent activity using your account’s data.

Anthropod’s public MCP access uses the mcp:read scope. It lets your assistant retrieve analytics, inspect configuration, read reports, and export existing data. Export tools create downloadable files; they do not change your conversations, customers, or configuration.

Connect to Anthropod

Setting	Value
Server URL	`https://mcp.anthropod.ai/mcp`
Transport	Streamable HTTP
Access	Account-scoped read access (`mcp:read`)
Credential management	Dashboard → Developer Hub → Configuration → MCP Access

Open MCP Access in the dashboard and confirm access is enabled for your account. Contact support if it needs to be enabled.
Add the server URL to your assistant’s MCP connector settings.
Complete the OAuth authorization flow offered by your client. For clients that require MCP client credentials, use the MCP client ID and secret from Developer Hub.
Ask your assistant to list the available services and read the Anthropod MCP Guides and Starter Templates resource before starting a broad analysis.

Setup guides

Follow the detailed setup guide for your assistant, then use the starter templates to begin exploring your data.

Assistant	Setup guide
Claude	Claude Web App setup
ChatGPT	ChatGPT Web App setup

MCP credentials are managed separately from the REST API key. Keep client secrets in your client’s secure credential settings.

Read capabilities

Available data depends on your account, connected providers, and configured analytics modules. The connected server’s tool list describes the inputs for each tool.

Capability	Tools
Discover services and project context	`get_project_services`, `get_mcp_server_context`
Inspect analytics filters and module configuration	`get_analytics_filters`, `get_module_configuration`
Review agent performance and QA	`get_agent_feedback_analytics`
Analyze customer insights, concerns, and queries	`get_customer_insights_analytics`, `get_concern_query_analytics`
Read conversations and customer journeys	`get_conversations`, `get_customers`
Read saved reports and their results	`list_custom_reports`, `get_custom_report_definition`, `get_custom_report_output`
Inspect Voice Agents and available filters	`list_voice_agents`, `get_voice_agent_configuration`, `get_voice_agent_filters`
Review Voice Agent activity and analytics	`get_voice_agent_home_analytics`
Read Voice Agent conversations	`get_voice_agent_conversations`, `get_voice_agent_conversation`
Read Voice Agent customer journeys	`get_voice_agent_customers`, `get_voice_agent_customer`
Export conversation and customer analytics	`create_analytics_export`, `get_analytics_export_status`
Export Voice Agent data	`create_voice_agent_export`, `get_voice_agent_export_status`

Guides and starter templates

The MCP server includes resources your assistant can discover and read after connecting:

Resource	What it contains
Authentication and credential rotation	Access-token setup, renewal, and credential rotation through Developer Hub
Anthropod MCP Guides and Starter Templates	Analysis guidance and starter templates for agent insights, process insights, marketing and customer insights, exploration, and reporting

To load the analysis guide, ask:

Read the "Anthropod MCP Guides and Starter Templates" resource.
List the services available to my account and help me choose a starter
analysis for my service and date range.

For MCP clients that expose resource identifiers directly, the guide is available at:

anthropod-mcp://guides/analytics-and-starter-templates

Example prompts

Open a category in the starter-template document, then replace the bracketed values with your service and dates. The templates work with Claude and ChatGPT after connecting Anthropod MCP.

Category	What you can explore	Document link
Agent performance	QA, training calls, agent comparisons, and coaching priorities	Agent performance prompts
Process insights	Funnel leakage, response times, follow-ups, and missed demand	Process insight prompts
Marketing and customer insights	Lead quality, campaigns, customer segments, and conversion outcomes	Marketing and customer prompts
Exploration	Changes in outcomes, growth opportunities, and root-cause analysis	Exploratory prompts
Reporting	Executive reviews, agent QA workbooks, and customer insight reports	Reporting prompts

For broader investigations, agree on the service, date range, outcome definition, and comparison period before starting the analysis.

MCP or REST API?

Interface	Use it for
MCP	Asking an AI assistant to explore data, explain results, and prepare analyses
REST API	Connecting application backends, CRM integrations, and data pipelines directly over HTTP