About Cinderfi
Cinderfi is an MCP server for tax-aware retirement planning in Canada and the US. It provides individual and couples projections, benefit timing, tax analysis, simulations, and withdrawal strategies for agents and compatible clients.
Key features
- 01Retirement projections for individuals and couples
- 02CPP/OAS and Social Security timing
- 03Canadian and US tax calculations
- 04Monte Carlo simulations
- 05Withdrawal optimization and historical backtesting
Use cases
- 01Compare retirement scenarios for a household
- 02Assess when to claim public retirement benefits
- 03Evaluate RRSP meltdown or Roth conversion strategies
MCP support
Exposes 19 retirement-planning tools to MCP clients through Streamable HTTP and SSE. Clients connect to its endpoint using an API key or supported auth registration flow.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://mcp.cinderfi.com/mcpclaude mcp add --transport http cinderfi https://mcp.cinderfi.com/mcp{"mcpServers":{"cinderfi":{"type":"http","url":"https://mcp.cinderfi.com/mcp"}}}Open in Cursor{"servers":{"cinderfi":{"type":"http","url":"https://mcp.cinderfi.com/mcp"}}}Open in VS Code{"mcpServers":{"cinderfi":{"type":"http","url":"https://mcp.cinderfi.com/mcp"}}}{"mcpServers":{"cinderfi":{"serverUrl":"https://mcp.cinderfi.com/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://mcp.cinderfi.com/mcpAuth requiredTools
Details
From the project READMEcinder.fi MCP
Tax-aware retirement planning for Canada and the US, available as a Claude Code plugin, an Agent Skills skill, and a Model Context Protocol server.
Compatible with Claude Code, Cursor, VS Code Copilot, OpenAI Codex, Gemini CLI, Goose, and any other MCP- or Agent Skills-compatible agent.
What it does
19 tools covering:
- Retirement projections (single + couples)
- CPP / OAS / Social Security timing
- Tax calculations (all Canadian provinces + US states)
- GIS eligibility and clawback analysis
- RRSP meltdown / Roth conversion strategy
- Monte Carlo simulation
- Withdrawal order optimization
- Historical backtesting (Shiller 1871–2024)
- Purchase and windfall impact simulators
- Scenario comparison
- Personalized retirement insights
Claude Code plugin
The fastest path in Claude Code — installs the MCP server plus a retirement-planning skill that sequences the tools:
/plugin marketplace add cinderfi/cinderfi-mcp
/plugin install cinderfi@cinderfi
Then set your key so the server authenticates:
export CINDERFI_MCP_KEY=your_api_key # from https://cinder.fi/mcp
Agents that support the auth.md protocol can also register for their own credential through a user-approved claim ceremony — no pasted key required.
Get started (any client)
- Get a free API key at cinder.fi/mcp
- Connect your agent to
https://mcp.cinder.fi/mcp
Claude Desktop
{
"mcpServers": {
"cinderfi": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://mcp.cinder.fi/mcp"],
"env": {
"AUTHORIZATION": "Bearer YOUR_API_KEY"
}
}
}
}
Any MCP-compatible client
- Endpoint:
https://mcp.cinder.fi/mcp - Auth:
Authorization: Bearer YOUR_API_KEY(or the auth.md registration flow) - Transports: Streamable HTTP, SSE
Pricing
| Free | Pro | |
|---|---|---|
| Calls/day | 5 | Unlimited |
| Single-person tools | ✅ | ✅ |
| Couples tools | — | ✅ |
| Monte Carlo | — | ✅ |
| Optimizer | — | ✅ |
| Backtest | — | ✅ |
| Scenario comparison | — | ✅ |
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.