Skip to main content
MCP Integration

Connect MCP clientsto AskElephantcontext

AskElephant's MCP server gives Cursor, Claude, VS Code, and Windsurf access to product context, brand voice guidance, and customer ROI demo data.

Get startedView docs
Standby
local stdio server
CClaude
WWindsurf
CuCursor
VSVS Code
MCP
Protocol Bridge
Revenue OS
Brand Voice
Customer ROI

Works with MCP clients that support local stdio servers

Claude
VS Code
Cursor
Windsurf
Codex

QUICK START

One config. Three AskElephant tools.

Register the local AskElephant MCP server with a client that supports stdio transport. The current implementation runs from this repo with a command-based config.

Claude Settings > MCP Connectors
// Add this to your MCP client config
{
"mcpServers": {
"askelephant": {
"command": "uv",
"args": ["--directory", "/path/to/www.askelephant.ai/mcp-server", "run", "server.py"]
}
}
}
✓ Connected — 3 demo tools available locally

MCP TOOLS

What can AI access through AskElephant MCP?

Three purpose-built tools that give compatible AI clients structured AskElephant context for demos, testing, and internal workflows.

Revenue OS Features

Access structured product information including AI agent capabilities, boundaries, and performance metrics for CRM automation, coaching, alerts, handoffs, and chat.

  • Query agent capabilities and boundaries
  • Access performance metrics per agent
  • Filter by specific agent or get all
  • Platform capability inventory
> get_revenue_os_features(agent_id: "crm-automation")

Brand Voice Guidelines

Retrieve messaging guidelines, voice attributes, approved phrases, and response templates aligned with AskElephant's 'Making Work Human Again' philosophy.

  • Empathy-first voice attributes
  • Approved messaging frameworks
  • Response templates for common scenarios
  • Key phrases and positioning language
> get_brand_voice_guidelines(section: "messaging_frameworks")

Customer ROI Metrics

Access reference customer success metrics, case studies with proof points, and industry benchmarks to support demos, sales conversations, and content creation.

  • Aggregate ROI metrics across customers
  • Detailed case studies with quotes
  • Industry-specific benchmarks
  • Reference proof points for demos
> get_customer_roi_metrics(company: "PestShare")

HOW IT WORKS

How do you set up AskElephant MCP?

Four steps from local setup to structured answers. No custom adapters required.

1

Install the MCP Server

Install the AskElephant MCP server locally from this repo. The current implementation runs over stdio transport.

2

Connect Your AI Client

Register the local server in Cursor, VS Code, Claude, Windsurf, or another MCP client that supports stdio.

3

Query Your Revenue Data

Ask natural language questions about customers, product features, brand guidelines, and ROI metrics.

4

Get Structured Answers

Receive structured responses from AskElephant's MCP tools using the server's current demo data set.

WHY MCP

The USB-C for AI — one standard to connect everything

Without MCP, connecting 3 AI tools to 5 data sources requires 15 custom integrations. With MCP, you need just 8 connectors — 3 clients + 5 servers. Build once, integrate everywhere.

15
Custom integrations
Without MCP
8
MCP connectors
With MCP
97M+
Monthly SDK downloads
75%
Companies adopting MCP

FAQ

Frequently asked questions about MCP

What is MCP and how does it work with AskElephant?
MCP (Model Context Protocol) is an open standard for connecting AI applications to external data sources. AskElephant's MCP server exposes three tools that give AI assistants access to structured product information, brand guidelines, and customer success examples for demos and testing.
Which AI clients support MCP?
MCP is supported by Cursor, VS Code, Claude, Windsurf, and other compatible clients. Exact setup depends on which transports a client supports, and the server in this repo currently runs locally over stdio.
Is my data secure when using MCP?
Yes. MCP requests are sandboxed behind the server, which enforces business rules and access controls. Every request, input, and output is logged for auditability. The server runs locally via stdio transport — your data never passes through a third-party proxy.
How does MCP differ from a regular API?
While traditional APIs require custom integration code for each AI tool, MCP provides a universal standard — like USB-C for AI. Write one MCP server, and every compatible AI client can access it. No custom adapters, no M×N integration problem.
Can non-technical team members use MCP?
Once configured, MCP is invisible to end users. Your sales and CS teams simply ask their AI assistant questions in natural language, and the assistant automatically queries AskElephant's MCP server for structured reference data. No code or technical knowledge required.
How do I measure ROI from MCP integration?
Track three metrics: time saved on manual data lookups, accuracy improvement in AI-generated content (fewer hallucinations), and adoption rate across your team's AI tools. Customers typically see 50+ hours/month in time savings from automated context retrieval.

Still not sure if AskElephant is right for you?

Let ChatGPT, Claude, Gemini, or Perplexity do the thinking for you. Click a button to see what your favorite AI says about AskElephant.

Ask ChatGPTAsk ClaudeAsk GeminiAsk Perplexity

These links open external AI services in a new tab. Login may be required.