Add Stack-Based MCP Server Recommendations Based on Project Files

[olearycrew]

Proposal Summary

When working on projects with specific tech stacks (identified by package.json, docker-compose.yml, openapi.yaml, README/DEVELOPER.md), users currently have to manually search for and install relevant MCP servers that would enhance their development workflow. This creates friction in the developer experience and users may miss out on valuable tools that would make their work more efficient.

Proposed Solution

Implement an intelligent recommendation system that automatically suggests relevant MCP servers based on the detected project stack files. The feature would:

1. Analyze project files like package.json, docker-compose.yml, openapi.yaml, and development documentation to determine the technology stack
2. Match the detected stack components with relevant MCP servers from the marketplace
3. Present these recommendations in a user-friendly way similar to VS Code's "@recommendations" feature
4. Allow users to easily install recommended MCP servers with a single click

Additional context

Alternatives Considered

1. Manual documentation of "recommended MCP servers" for various tech stacks, but this would require constant updates and wouldn't be contextual to the user's specific project
2. Using a fixed set of "popular" MCP servers regardless of project type, but this would not provide personalized recommendations
3. Building an AI-based suggestion system that requires cloud processing, but this would add unnecessary complexity and potential privacy concerns

User Impact

This feature would benefit all Kilo Code users by:

• Reducing the learning curve for new users unfamiliar with available MCP servers
• Saving time for experienced users when setting up new projects
• Improving discovery of specialized MCP servers that match project requirements
• Enhancing the overall developer experience with contextual tooling

Implementation Thoughts

The implementation could follow these steps:

1. Create a detection system for common project files and patterns:

   • Parse package.json for Node.js dependencies and project type
   • Identify Docker components from docker-compose.yml
   • Detect API specifications from openapi.yaml
   • Scan README/DEVELOPER.md for technology keywords

2. Build a mapping database that connects detected technologies to relevant MCP servers:

   • For React projects, suggest React-specific tools
   • For Docker-based projects, recommend container management tools
   • For API projects, suggest API testing and documentation servers

3. Develop a recommendation UI similar to VS Code's extension recommendations:

   • Show in the MCP Server Marketplace view
   • Include an explanation of why each server is recommended
   • Allow filtering/sorting of recommendations
   • Provide one-click installation

4. Add telemetry to track recommendation effectiveness (with user consent)

Related Features

• VS Code's built-in extension recommendations based on file types
• Kilo Code's existing MCP Server Marketplace integration
• The "Recommended Extensions" feature in VS Code that suggests extensions from .vscode/extensions.json

Contribution Path

This feature would be ideal for community contributors familiar with VS Code extension development as it requires:

• Knowledge of file format parsing
• Understanding of VS Code's extension API
• Familiarity with MCP server capabilities