Kansei Mcp Server
Research which SaaS tools fit your stack using five unified MCP tools instead of dumping hundreds of vendor pages into context.
Overview
io.github.kansei-link/kansei-mcp-server is an MCP server for the Idea phase that delivers SaaS intelligence across 1,000+ services via 5 unified tools.
What is this MCP server?
- 5 unified MCP tools spanning SaaS intelligence across 1,000+ services
- Claims 91–96% token savings versus pasting large vendor corpora into the model
- npm package @kansei-link/mcp-server plus remote streamable-http endpoint on Railway
- stdio and hosted remote transport options for different MCP client setups
- Version 1.0.0 oriented at agent-side vendor discovery not single-API wiring
- 1,000+ services coverage claimed
- 91–96% token savings claimed
Community signal: 1 GitHub stars.
What problem does it solve?
Stack research balloons context with vendor pages and the agent still picks the wrong tool for your constraints.
Who is it for?
Indie builders comparing third-party SaaS categories before they commit architecture or buy subscriptions.
Skip if: Teams that already finalized their stack and only need one known API’s OpenAPI client.
What do I get? / Deliverables
You query structured SaaS intelligence through MCP with far smaller prompts while shortlisting integrations.
- Shortlisted SaaS candidates grounded in MCP tool output
- Lower-token research sessions versus raw web dumps
- Documented integration choices to carry into Validate and Build
Recommended MCP Servers
Journey fit
Choosing integrations and positioning happens early when you compare categories, pricing, and alternatives before committing to build. research captures vendor landscape and capability mapping—the server markets coverage of 1,000+ services for that job.
How it compares
SaaS discovery MCP with token-efficient tools, not a single-vendor integration server.
Common Questions / FAQ
Who is io.github.kansei-link/kansei-mcp-server for?
Solo builders and agent users researching SaaS options across many categories without massive copy-paste research prompts.
When should I use io.github.kansei-link/kansei-mcp-server?
Use it during Idea research (and early Validate scoping) when you compare services, features, and fit before Build integrations.
How do I add io.github.kansei-link/kansei-mcp-server to my agent?
Install @kansei-link/mcp-server for stdio, or configure the streamable-http remote URL https://kansei-link-mcp-production-b054.up.railway.app/mcp if your client supports hosted MCP.