Using Responsive MCP server with Generative AI Tools

Overview

What is MCP?

Model Context Protocol (MCP) is a standard that allows generative AI tools (like chat-based assistants) to securely access external data sources and systems. Instead of relying only on what the AI already knows, MCP lets it retrieve and use real-time, trusted information.

What Responsive MCP Enables

Responsive’s MCP server connects your AI tools directly to your organization’s approved RFP content and knowledge base. Instead of copying answers from Responsive into AI, MCP lets AI work directly with your approved content. This allows you to:

• Retrieve accurate, approved answers from your content library

• Generate responses grounded in your company’s data

• Create projects and comments in your preferred AI tools that will appear in Responsive

• Improve consistency and reduce manual effort in proposal workflows

In short, it brings trusted RFP intelligence into your AI workflow.

Prerequisites

Before connecting, ensure the following:

Access Requirements

Responsive AI feature is enabled on Responsive.

You have Administrator permissions on the AI tool you are using.

Supported MCP-Compatible Tools

Responsive MCP server can be used with AI tools that support MCP integrations.
For example:

• ChatGPT

• Claude

• Copilot

Connecting Responsive MCP to your AI tool

In your AI tool, navigate to MCP or integration settings.

Create a new MCP server connection with the following MCP URL:

https://app.rfpio.com/oa/v1/mcp

Note: In the example above, app is the default domain. If you are using a different or custom domain, use that one instead.

When done, you'll be asked to login to your Responsive account.

Once completed, you should be able to access Responsive data and return relevant results from the AI tool.

Using Responsive MCP in your AI tool

Based on the AI platform you are using, ensure you have added Responsive as a data source or enabled the created Responsive connector.

Once connected, users can start interacting with Responsive data directly within their AI tool.

Instead of asking generic questions, you can prompt the AI to:

• Use your organization’s knowledge

• Retrieve past responses

• Generate answers grounded in approved content

• Create projects in your AI tools that will appear in Responsive

The AI will automatically call the appropriate MCP tools/actions behind the scenes.

Available MCP Tools/Actions

Responsive MCP provides a set of tools that your AI tool can use to retrieve information and support RFP workflows. You do not need to invoke these tools by name, but it helps to understand what each one does and when it is useful.

Find and understand your project

get_project_list

What it does:
Returns a list of available projects.

When to use it:
Use this when you want to identify which project to work on.

Example prompt:

List the projects that I’m a part of.

get_project_details

What it does:
Retrieves details for a specific project.

When to use it:
Use this when you need project-level context before answering questions or reviewing progress.

Example prompt:

Get the details for the Acme Security Questionnaire project so I can understand its status, timeline, and context.

get_project_sections

What it does:
Returns the sections within a project.

When to use it:
Use this when you want to understand project structure or work section by section.

Example prompt:

Show the section list and progress for ACME Corp.

get_project_question

What it does:
Retrieves a specific question from a project.

When to use it:
Use this when you want to inspect, review, or prepare to answer a particular question.

Example prompt:

Get the question in the Acme project about disaster recovery so I can draft a stronger answer.

Identify what needs attention

get_unanswered_questions

What it does:
Finds unanswered questions in a project.

When to use it:
Use this when you want to identify response gaps and prioritize remaining work.

Example prompt:

Show my unanswered questions for SRM RFP.

Find and use content

search

What it does:
Searches Responsive content to find relevant answers, source material, or prior knowledge.

When to use it:
Use this when you want to locate approved content before drafting or reviewing a response.

Example prompt:

Do we have an ISO 27001 certificate? Share the source.

fetch

What it does:
Retrieves specific content or details from Responsive for use in the conversation.

When to use it:
Use this when you want the AI tool to pull in relevant information before answering or drafting.

Example prompt:

Fetch the current answer content we have for encryption, access controls, and audit logs so I can review it before drafting a response.

Draft responses

generate_draft_response

What it does:
Generates a draft response to a question using available Responsive content and project context.

When to use it:
Use this when you want a first draft for an RFP question instead of writing from scratch.

Example prompt:

Generate a response for: ‘security policies’

Understanding user context

get_my_profile

What it does:
Returns information about the current Responsive user profile.

When to use it:
Use this when the AI needs to understand who is signed in or tailor work based on the current user.

Example prompt:

Get my profile so you can confirm which Responsive account and workspace this session is using.

Create a project

create_project

What it does:
Allows you to use conversational language to create a project that will appear in Responsive without leaving the AI tool.

When to use it:
Use this when you want to create a project. Enter the details for the required and optional fields as part of your prompt, including:

• Project name

• Project type

• Primary contact

• Due date

• Client details (optional)

• Project description (optional)

• Project stage

You can include document files and organization-specific project fields.

Example prompt:

Create a new Responsive project called Anderson Manufacturing. The project is to respond to an RFP from that company which is due August 31, 2026, for $50,000. The primary contact is Chase Smith. Include the uploaded file.

Create a comment

create_comment

What it does:
Allows you to use conversational language to create comments and associate them with a section and question.

When to use it:
When you want to create a comment for a project in Responsive without leaving the AI tool.

To create a comment, enter the comment in the chat window. If you do not include a section or question, the comment will be added at the project level.

Note: You must have permission to add comments to the project, and any unauthorized users cannot add comments.

Example prompts:

Create a comment for the Anderson Manufacturing project. @mention Chase Smith and ask him to make sure the answers in the Security section are accurate, especially the question about the type of encryption we use. 

Get the questions that need my attention or my review, then create a comment for the first unanswered question. The comment should say to make sure it is finished by August 1.