@profullstack/mcp-server

# MCP Server Implementation TODOs This document outlines the remaining tasks needed to fully implement the MCP standard methods as defined in the [MCP Standard Methods](docs/mcp_standard_methods.md) document. ## Inference Support - [ ] Enhance the inference implementation in `modelManager.js` to support more advanced inference capabilities: - [ ] Add support for streaming responses - [ ] Implement proper error handling with standardized error codes - [ ] Add support for different inference parameters (temperature, max_tokens, etc.) - [ ] Implement rate limiting for inference requests ## Tools and Resources - [ ] Implement the tools execution endpoint: - [ ] Create a proper implementation for `/tools/:toolName` endpoint - [ ] Add a mechanism for modules to register tools - [ ] Implement tool execution logic - [ ] Implement the resources access endpoint: - [ ] Create a proper implementation for `/resources/:resourceUri` endpoint - [ ] Add a mechanism for modules to register resources - [ ] Implement resource access logic ## Module Management - [ ] Implement module enable/disable functionality: - [ ] Add `/modules/:moduleId/enable` endpoint - [ ] Add `/modules/:moduleId/disable` endpoint - [ ] Update the module loader to respect enabled/disabled status ## Testing - [ ] Write unit tests for inference functionality: - [ ] Test streaming responses - [ ] Test error handling - [ ] Test parameter validation - [ ] Write unit tests for tools and resources: - [ ] Test tool registration - [ ] Test tool execution - [ ] Test resource registration - [ ] Test resource access - [ ] Write unit tests for module enable/disable functionality ## Pre-commit Hook - [ ] Update the pre-commit hook to run tests: ```sh #!/usr/bin/env sh . "$(dirname -- "$0")/_/husky.sh" # Run lint-staged to lint and format staged files pnpm lint-staged # Run tests before commit echo "Running tests before commit..." pnpm test || exit 1 echo "Pre-commit checks passed. Proceeding with commit." ``` ## Documentation - [ ] Update documentation to reflect the new functionality - [ ] Add examples for using the inference API - [ ] Add examples for creating and using tools - [ ] Add examples for creating and accessing resources