Sunwise Model Context Protocol (MCP)
Check out more about Model Context Protocol (MCP) to enable AI agents to securely access and interact with your Sunwise information.
How Sunwise MCP Works
Sunwise MCP (Model Context Protocol) acts as a bridge between AI assistants and the Sunwise platform. Instead of the AI guessing or simulating data, MCP allows it to securely read and act on real Sunwise information using clearly defined tools.
At a high level, this is how it works:
-
You request something in natural language
For example: creating a project, finding a contact, or generating a proposal. -
The AI understands the intent
The assistant translates your request into a structured action. -
MCP selects the correct Sunwise tool
Each action (creating contacts, projects, proposals, reports, etc.) has a dedicated tool with strict rules. -
Sunwise executes the action
The request is validated, executed inside Sunwise, and real data is returned. -
The AI responds with accurate results
Since the data comes directly from Sunwise, responses are always up to date and reliable.
In simple terms, MCP lets the AI work with Sunwise, not just talk about it.
Benefits of Using MCP
Using Sunwise MCP provides several important advantages, especially for teams automating workflows or building AI-powered experiences.
-
No hallucinations
The AI can only use what Sunwise actually provides, reducing errors and made-up data. -
Secure by design
Access is controlled by authentication and permissions already defined in Sunwise. -
Structured and predictable
Each tool has clear inputs and outputs, making integrations easier to maintain. -
Real actions, not suggestions
MCP doesn’t just recommend steps — it can create contacts, projects, proposals, and reports. -
Faster workflows
Repetitive tasks can be handled directly through conversational interactions. -
Scalable automation
The same tools can be used by humans, bots, or external systems consistently.
Available Tools
Sunwise MCP exposes a set of tools that represent core actions in the Sunwise platform. These tools are grouped by purpose and are designed to be easy to compose into workflows.
Contact Management
- Create new contacts
- Update existing contacts
- List and search contacts
- Retrieve contact details by ID
- Access projects linked to a contact
Project Management
- Create projects (with or without consumption data)
- Retrieve recent projects
- Get projects associated with a specific contact
- Access project summaries
Proposal Management
- Create commercial proposals inside projects
- Generate proposals from presets
- Retrieve proposals by project or ID
- Search proposals by name
- Mark and retrieve favorite proposals
Financing
- Retrieve financing options for a proposal
- Create financing plans
- Get detailed financing information for proposals
Reports & Summaries
- Retrieve active reports with public URLs
- Access detailed report data
- List and retrieve project summaries
- List and retrieve proposal summaries
Each tool is purpose-built, meaning the AI can only perform exactly what the tool allows — nothing more, nothing less. This keeps interactions safe, reliable, and aligned with Sunwise business logic.
Setting things up
Authentication Methods
The Sunwise MCP server supports the OAuth Flow authentication method, meaning you can access to it by login into it with your Sunwise credentials.
Quick Guide: Add Sunwise MCP to ChatGPT
The exact steps depend on the ChatGPT client you are using (web, desktop, enterprise), but the overall flow is the same.
- Open Settings in ChatGPT.
- Navigate to Apps & Connectors.
- Look for the Advance settings and activate Developer mode
- Go back and select "create" to create a connector.
- Configure the server:
- Server URL:
https://production.sunwise.ai/boty/mcp - Authentication: You can do a normal OAuth login with your credentials.
- Save the configuration.
- Test the connection with a simple read action (for example, listing contacts or retrieving recent projects).
Other LLM Provider Setup Guides
Each AI provider has specific setup instructions for MCP servers. Please consult the official documentation for your provider:
- Claude Desktop: MCP setup documentation
- Claude Code: MCP setup documentation
- OpenAI: MCP integration guide
- Claude.ai: Go to settings > Integrations > + Add integration, then use
https://production.sunwise.ai/boty/mcp - Cursor: MCP configuration guide
- Windsurf: MCP setup instructions
- VS Code: MCP integration docs