AppFrame MCP Server Setup

AppFrame MCP Server - Setup Guide

Status: Checking...

User: Not logged in

Mode: N/A

Protocol: MCP 2024-11-05

The AppFrame MCP Server enables AI assistants to access PIMS database functionality while running as your authenticated appFrame user. This allows LLMs to query data, explore schema, and assist with database tasks.

https://vetslivestockmanagement.com/Application/AI/MCP/MCPHandler.ashx

Client Configuration

MCP SuperAssistant

Manual JSON

Edge Copilot

Gemini/Chrome

MCP SuperAssistant Chrome Extension

Note: MCP SuperAssistant acts as a bridge between browser-based LLMs and MCP servers. It allows ChatGPT, Claude, Gemini and other web LLMs to use MCP tools.

Install MCP SuperAssistant from the Chrome Web Store
Click the extension icon and go to Settings
Click Add Server and select Streamable HTTP
Enter the configuration below:

Name: AppFrame-PIMS URL: https://vetslivestockmanagement.com/Application/AI/MCP/MCPHandler.ashx Type: Streamable HTTP

Click Save and ensure the server shows as Connected
Open ChatGPT, Claude, or another supported LLM website
The extension will inject MCP tools - you can now ask the LLM to query PIMS!

Important: You must be logged into appFrame in the same browser session. The MCP server uses your appFrame session for authentication.

Manual Configuration (JSON)

For MCP clients that use JSON configuration files:

{ "mcpServers": { "appframe-pims": { "type": "streamableHttp", "url": "https://vetslivestockmanagement.com/Application/AI/MCP/MCPHandler.ashx", "description": "AppFrame PIMS database access via MCP" } } }

For stdio-style clients (requires proxy)

{ "mcpServers": { "appframe-pims": { "command": "npx", "args": [ "mcp-remote", "https://vetslivestockmanagement.com/Application/AI/MCP/MCPHandler.ashx" ] } } }

Microsoft Edge Copilot

Current Status: Edge Copilot does not yet natively support custom MCP servers. Use MCP SuperAssistant as a bridge to enable MCP tools in Copilot.

Workaround: MCP SuperAssistant Bridge

Install MCP SuperAssistant in Edge (it's compatible)
Configure the AppFrame MCP server (see SuperAssistant tab)
When using Copilot, the extension will provide tool access

Future: Native Copilot Integration

Microsoft is developing native MCP support for Copilot. When available, configuration may look like:

# Copilot Settings > Extensions > MCP Servers Add Server: Name: AppFrame-PIMS Endpoint: https://vetslivestockmanagement.com/Application/AI/MCP/MCPHandler.ashx Auth: Browser Session (same origin)

Google Gemini / Chrome AI

Current Status: Gemini web interface does not yet support custom MCP servers directly. Use MCP SuperAssistant as a bridge.

For Gemini CLI (command line)

gemini mcp add appframe-pims https://vetslivestockmanagement.com/Application/AI/MCP/MCPHandler.ashx

Chrome Built-in AI (Experimental)

Chrome's built-in AI features don't yet support MCP. For now, use:

Install MCP SuperAssistant extension
Configure AppFrame MCP server
Access gemini.google.com while logged into appFrame

Available Tools

query_pims

Execute SQL queries against PIMS database. Use NOLOCK for SELECT queries. Use views (atbv_*, aviw_*) for proper permissions.

All Users (SELECT) Developer (DML)

get_table_info

Get detailed table/view schema including columns, data types, and primary keys.

All Users

list_tables

List tables and views in the database. Supports SQL LIKE pattern filtering.

All Users

list_stored_procedures

List stored procedures. Supports pattern filtering (e.g., 'astp_AI%').

All Users

get_developer_mode

Check your current permission level and what operations are allowed.

All Users

ping

Health check - returns server status, your username, and permission level.

All Users

ask_minion

Route questions to specialized V.E.T.S. Minion experts. Use 'auto' for intelligent routing or specify: Oz (AI/RAG), Florence (Medical), Penny (Billing), Lassie (Herds), Leggs (API/Debug), Otter (CRM), Badger (Auth), Baxter (Creation), Brick (Security), Clerk (Docs), Spot (Reports), Trigger (Workflow), Viper (Social), Dr. Dolittle (General).

All Users

report_conversation

Report IDE conversations to V.E.T.S. learning pipeline. Extracts learnings, tracks files/tables/procs touched, identifies RAG gaps. Call at end of productive sessions. Quality 6+ = productive session.

All Users

Permission Levels

Operation	Standard User	Developer	Master Dev	Admin
SELECT from Views (atbv_, aviw_)	[OK]	[OK]	[OK]	[OK]
SELECT from Tables (atbl_, stbl_)	[NO]	[OK]	[OK]	[OK]
INSERT / UPDATE / DELETE	[NO]	[OK]	[OK]	[OK]
CREATE / ALTER / DROP	[NO]	[NO]	[OK]	[OK]
Execute Stored Procedures	Whitelisted only	[OK]	[OK]	[OK]

Your Current Level: Please log in to appFrame to use MCP features.

Example Prompts

Once configured, try these prompts with your AI assistant:

"What tables are available for AI functionality? List all tables starting with atbl_AI" "Show me the schema for the atbl_AI_MinionsConversation table" "Query the top 10 most recent AI conversations with their token counts" "Check my current developer mode level and permissions" "List all stored procedures related to TeamDoc" "Ask the AI expert (Oz) how RAG works in V.E.T.S." "Ask Florence how to add a prescription for an animal" "Use ask_minion with auto routing: How do I create an invoice?"