The Salespeak MCP Server allows AI assistants to connect directly to Salespeak and query analytics about AI agent traffic, AI discovery behavior, and on-site AI chat conversations.
This makes it possible to ask natural-language questions like:
“Which AI agents visit my website most frequently, and what are they focusing on?”
“Have AI agents found my latest content updates from last week?”
“Summarize the top questions from AI chat conversations this month.”
The server is designed for Streamable HTTP MCP clients and supports OAuth 2.0 authentication.
Key Capabilities
1. Track AI Agent Activity & Intent
Use the Salespeak MCP server to understand which AI crawlers and agents are accessing your site, what pages they view, and what they appear to be trying to accomplish.
Example questions:
“Which AI agents visit my website most frequently, and what are they focusing on?”
“Break down AI agent traffic by agent type (ChatGPT, Perplexity, Claude, Gemini) and summarize their browsing behavior.”
“Which AI agents are visiting my product pages vs. blog content vs. documentation?”
Common outputs include:
agent volume trends
top visited pages
content category breakdowns
inferred browsing intent (research vs. shopping vs. comparison, etc.)
2. Monitor AI Discovery of New Content
Track whether AI agents are discovering your newest content and how they interpret it.
Example questions:
“Have AI agents found my latest content updates—like last week’s blog posts—and how are they interpreting them?”
“Which new pages were discovered by AI agents in the last 7 days, and what key topics did they extract?”
“Are AI agents misunderstanding or misclassifying any of my newly published pages?”
Common outputs include:
newly discovered URLs
extracted key topics
summarization accuracy signals
misclassification / mismatch alerts
3. Analyze AI Chat Conversations on Your Website
If you use Salespeak’s on-site AI chat assistant, you can analyze conversation trends and visitor intent.
Example questions:
“What are visitors asking my website AI chat assistant most often, and what themes are emerging?”
“Summarize the top questions from AI chat conversations this month and map them to buyer journey stages.”
Common outputs include:
top recurring questions
clustering of conversation themes
funnel stage mapping (awareness → consideration → decision)
conversion blockers and missing content insights
Getting Started
Requirements
To use the Salespeak MCP Server, you need:
A Salespeak account with access to analytics
An MCP-capable client (ChatGPT, Claude Desktop, Cursor, etc.)
OAuth2 credentials issued by Salespeak
Authentication (OAuth 2.0)
The Salespeak MCP server uses OAuth 2.0 to authenticate users and authorize access to their workspace analytics.
OAuth Flow Overview
The MCP client redirects the user to Salespeak’s authorization page.
The user approves access.
Salespeak issues an authorization code.
The MCP client exchanges the code for an access token.
The MCP client uses the access token to call MCP tools.
Token Behavior
Access tokens expire periodically
Refresh tokens may be supported depending on your client integration
Requests without a valid token return 401 Unauthorized
Transport: Streamable HTTP
Salespeak MCP is delivered via Streamable HTTP, meaning:
clients connect over standard HTTPS
responses may stream incrementally
tool calls are executed server-side with structured JSON output
This is ideal for analytics queries where large responses may stream progressively.
Tools Reference
The authenticated /mcp endpoint exposes 10 tools. All tools automatically resolve the caller's organization from the OAuth token. You cannot access data outside your authorized workspace.
Unless otherwise specified, all analytics tools return a structured response including a natural-language summary and key metrics. If a tool supports output_format, use "structured" for JSON-style results or "prose" for a narrative-only answer.
answer_prospect_question
Get sales-ready answers about products, pricing, objection handling, and competitive positioning from your Salespeak knowledge base.
| Parameter | Type | Required | Description |
|---|---|---|---|
query |
string | Required | The sales question or objection to address. |
session_id |
string | Optional | Session identifier for conversation continuity. |
tone |
string | Optional | Response style: "concise", "detailed", "executive". |
length |
string | Optional | Response length: "short", "long". |
persona |
string | Optional | Seller persona: "SDR", "AE", "VP Sales". |
deal_stage |
string | Optional | Deal stage context: "discovery", "demo", "negotiation". |
include_follow_ups |
boolean | Optional | Include suggested follow-up questions. Default: false. |
analyze_conversations
Analyze what visitors are asking in chat conversations. Returns session counts, top questions, intent breakdown, and geography.
| Parameter | Type | Required | Description |
|---|---|---|---|
query |
string | Required | Analytics question in natural language. |
time_range |
string | Optional | Time period: "1d", "7d", "30d", or "90d". Default: "7d". |
timezone |
string | Optional | Timezone for time-based grouping (e.g. "UTC", "America/New_York"). Default: "UTC". |
page |
string | Optional | Filter by landing page URL substring (e.g. "pricing"). |
intent |
string | Optional | Filter by intent classification (e.g. "high", "medium"). |
output_format |
string | Optional |
"structured" (JSON + summary) or "prose". Default: "structured". |
analyze_llm_traffic
Analyze LLM bot traffic and AI search engine performance. Understand how ChatGPT, Perplexity, Claude, Gemini, and others discover your content.
| Parameter | Type | Required | Description |
|---|---|---|---|
query |
string | Required | Analytics question in natural language. |
time_range |
string | Optional | Time period: "7d", "30d", or "90d". Default: "30d". |
timezone |
string | Optional | Timezone for time-based grouping (e.g. "UTC", "America/Los_Angeles"). Default: "UTC". |
analyze_traffic_performance
Analyze which pages and traffic sources drive high-intent conversations and demo bookings. Useful for identifying the best acquisition channels.
| Parameter | Type | Required | Description |
|---|---|---|---|
query |
string | Required | Analytics question in natural language. |
time_range |
string | Optional | Time period: "7d", "30d", or "90d". Default: "30d". |
timezone |
string | Optional | Timezone for time-based grouping (e.g. "UTC", "America/New_York"). Default: "UTC". |
source |
string | Optional | Filter by UTM source (e.g. "google", "linkedin"). |
page |
string | Optional | Filter by landing page URL substring. |
output_format |
string | Optional |
"structured" (JSON + summary) or "prose". Default: "structured". |
analyze_llm_citations
Analyze which knowledge base sources the AI cites most in conversations. Understand citation coverage and content gaps.
| Parameter | Type | Required | Description |
|---|---|---|---|
query |
string | Required | Analytics question in natural language. |
time_range |
string | Optional | Time period: "7d", "30d", or "90d". Default: "30d". |
timezone |
string | Optional | Timezone for time-based grouping (e.g. "UTC", "America/New_York"). Default: "UTC". |
topic |
string | Optional | Filter citations by topic (e.g. "pricing", "security"). |
output_format |
string | Optional |
"structured" (JSON + summary) or "prose". Default: "structured". |
detect_knowledge_gaps
Detect where the AI gets confused or lacks knowledge base citations. Find content gaps and areas needing improvement.
| Parameter | Type | Required | Description |
|---|---|---|---|
query |
string | Optional | Analytics question. Default: general gap analysis. |
time_range |
string | Optional | Time period: "7d", "30d", or "90d". Default: "30d". |
timezone |
string | Optional | Timezone for time-based grouping (e.g. "UTC", "America/New_York"). Default: "UTC". |
output_format |
string | Optional |
"structured" (JSON + summary) or "prose". Default: "structured". |
get_session_details
Get details of specific chat sessions. Retrieve full conversation history by session ID or by searching session content. If both are provided, session_id takes priority. If neither is provided, the tool returns an error.
| Parameter | Type | Required | Description |
|---|---|---|---|
session_id |
string | Optional | Specific session ID to retrieve. |
search_query |
string | Optional | Search term to find in conversations. |
limit |
integer | Optional | Maximum sessions to return. Default: 5. |
create_training_content
Add a question/answer pair to the knowledge base. Capture approved answers, objection handling, or FAQ entries.
| Parameter | Type | Required | Description |
|---|---|---|---|
question |
string | Required | The question or objection to address. |
answer |
string | Required | The approved answer/response. |
category |
string | Optional | Category: "pricing", "objections", "features", etc. |
submit_url_for_training
Submit a URL to be crawled and added to the knowledge base. Add documentation, blog posts, or product pages.
| Parameter | Type | Required | Description |
|---|---|---|---|
url |
string | Required | The URL to crawl and add. |
add_information_to_the_knowledge_bank
Add freeform text to the knowledge base. Content is processed by an LLM to extract Q&A pairs and automatically categorize information.
| Parameter | Type | Required | Description |
|---|---|---|---|
content |
string | Required | The information/text to process and add. |
title |
string | Optional | Title describing the content (helps organize extracted Q&As). |
custom_instructions |
string | Optional | Instructions for the extraction LLM (e.g. focus on pricing, security, competitors, etc.). |
Support
Need help?
Contact us at support@salespeak.ai. For faster debugging, include the MCP tool name, timestamp, and any error message or request ID from your client logs.
Visit salespeak.ai for product updates and documentation.