Quickchat AI MCP

What incentivai/quickchat-ai-mcp MCP server does

The incentivai/quickchat-ai-mcp MCP server connects a Quickchat AI Agent to applications that support the Model Context Protocol. It gives an AI application access to the agent's configured Knowledge Base, capabilities, and conversational behavior rather than exposing a fixed, narrowly defined utility.

This is useful when an organization has already configured an agent in Quickchat AI and wants another AI application to interact with it. The connected application can send messages to the Quickchat agent, while the Quickchat AI dashboard remains the place to manage the agent's knowledge, settings, MCP name, description, and optional command information.

The project documentation names Claude Desktop, Cursor, VS Code, and Windsurf as example client applications. Compatibility with other clients depends on their MCP support and configuration process.

How it works

After creating a Quickchat AI account and configuring an agent, activate MCP from the Quickchat AI application. The MCP setup supplies a scenario ID and an API key when authentication is enabled. An MCP-compatible client then starts the local package and passes those values as environment variables.

The incentivai/quickchat-ai-mcp MCP server uses the scenario ID to identify the Quickchat AI configuration. The API key protects access when the Quickchat MCP integration requires one. Quickchat's MCP page provides the client configuration snippet with the relevant MCP name and values.

The interaction is not limited to a preselected static tool. Instead, the client can communicate with the configured Quickchat AI Agent through its conversational interface. Conversations can be viewed in Quickchat Inbox, although the recorded transcript represents the interaction between the two AI systems and may not exactly match the end user's original messages.

Setup and configuration

Create a Quickchat AI account, start the available trial, and configure the agent's Knowledge Base, capabilities, and settings. Activate MCP in the Quickchat AI dashboard, then provide its name, description, and optional command. These details help client applications determine when the MCP should be used.

Install uv first. For Claude Desktop or Cursor, run the package with uvx using the package name quickchat-ai-mcp. The client configuration includes:

• SCENARIO_ID: the Quickchat AI scenario identifier.
• API_KEY: the Quickchat AI API key when the MCP integration requires authentication.

To share the MCP with other users without exposing a secret, disable the Quickchat MCP page's “Require API key” option and distribute a configuration containing only the scenario ID. The documentation explicitly warns against publishing the Quickchat API key.

Tools and capabilities

The incentivai/quickchat-ai-mcp MCP server provides:

• Access to the configured Quickchat AI Knowledge Base.
• Conversation with a Quickchat AI Agent through an MCP-compatible client.
• Agent behavior governed by Quickchat AI capabilities and settings.
• MCP metadata management through the Quickchat AI dashboard.
• Conversation visibility through Quickchat Inbox.

Limitations and notes

The README does not define a separate list of named MCP tools; its main capability is open-ended messaging with the configured agent. Client setup varies outside the documented Claude Desktop and Cursor examples. Public configurations can omit the API key only after the Quickchat MCP integration is configured not to require one. The repository does not state a software license or provide standalone service pricing details.