Behavioral Audit MCP
Checks draft support replies for missing policy disclosures that the customer's situation makes relevant, returning a pass/gap verdict with suggestions.
README
Behavioral Audit MCP
Check a draft support reply for what it should say and doesn't, before it goes to the customer.
Part of Agentic Diaries. One hosted tool, disclosure_check.
What it does
Give it what a customer said and the reply you are about to send. It flags any should-know policy the customer's situation made relevant that your draft failed to proactively surface: a refund window, an auto-renewal, a retention offer, a warranty term. The knew-but-did-not-say gap, on a single message.
It returns a verdict, pass or gap. On a gap, it names each missed policy and the line that should have been there. It detects and suggests. It does not rewrite your reply, send anything, or keep the message beyond the check.
Install
One line. No login, no signup.
Claude Code:
claude mcp add --transport http behavioral-audit https://receipts-mcp.vercel.app/api/mcp
If you add it mid-session, restart Claude Code (or run /mcp) so the new tool appears.
Claude Desktop (or any MCP client), add to your config:
{
"mcpServers": {
"behavioral-audit": {
"type": "http",
"url": "https://receipts-mcp.vercel.app/api/mcp"
}
}
}
The tool: disclosure_check
Input:
customer_message(string) , what the customer said, verbatim.draft_reply(string) , the reply you are about to send.
Output: a verdict plus, on a gap, the policies your draft left out.
Example
A customer asks to cancel. Your draft confirms it but never mentions the refund window they still qualify for.
Request:
{
"customer_message": "I want to cancel my subscription today.",
"draft_reply": "Done, your subscription is cancelled. Sorry to see you go."
}
Result:
{
"verdict": "gap",
"misses": [
{
"id": "refund_window",
"title": "Refund window",
"suggestion": "You are within your 14-day refund window, so you can still request a refund for this billing period.",
"reason": "The customer is cancelling and qualifies for a refund they were not told about."
}
]
}
A clean reply returns verdict: "pass" with an empty misses list. That means clear to send, not that nothing ran.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.