Cowboy Model Finder
Wire a remote MCP product-recommendation wizard into your agent when prototyping guided-buying flows for physical goods.
Overview
Cowboy Model Finder is a MCP server for the Idea phase that runs a hosted interactive e-bike wizard with AI recommendations over streamable HTTP or SSE.
What is this MCP server?
- Hosted remote MCP at Heroku with streamable-http and SSE endpoints—no local npm install in server.json
- Interactive wizard flow for Cowboy e-bike model selection with AI recommendations
- Version 1.0.0 published against MCP server schema 2025-12-11
- Official Cowboy brand integration (repository cowboyhq/model-finder-mcp-app, website cowboy.com)
- Reference implementation for branded MCP apps that combine UX wizards with agent tool calls
- Server version 1.0.0
- 2 remote transports: streamable-http and SSE
- 0 local packages in server.json—fully remote deployment
What problem does it solve?
Agents cannot natively run a branded, step-by-step product finder unless you expose that flow through a protocol they already understand.
Who is it for?
Builders prototyping conversational commerce, remote-hosted MCP demos, or studying dual-transport (HTTP + SSE) integrations without shipping their own backend first.
Skip if: Teams that need a generic bike database, self-hosted stdio tooling, or non-Cowboy product catalogs.
What do I get? / Deliverables
After you register the remote MCP URL, your agent can invoke the Cowboy recommendation wizard instead of hallucinating model specs.
- Agent-callable Cowboy model-finder tools over hosted MCP
- Working reference for dual remote transport configuration
- End-user-oriented recommendation flow without custom backend code
Recommended MCP Servers
Journey fit
Canonical shelf is early journey discovery because the server’s purpose is helping end users explore and narrow product options before purchase. Discover fits interactive exploration and AI-guided model matching rather than checkout, ads, or post-launch ops.
How it compares
Remote branded recommendation MCP, not a local npm integration skill for your own SKU data.
Common Questions / FAQ
Who is io.github.cowboyhq/cowboy-model-finder for?
Solo builders and small teams exploring hosted MCP patterns for guided product discovery and AI-assisted recommendations in e-commerce agents.
When should I use io.github.cowboyhq/cowboy-model-finder?
Use it during Idea and discovery work when you want a live reference for wizard-style MCP flows or when building agent experiences that should defer to Cowboy’s official finder.
How do I add io.github.cowboyhq/cowboy-model-finder to my agent?
Add the published remote MCP URL (streamable-http or SSE) from the server manifest to your client’s MCP remote servers list—no API key is listed in server.json.