A demo MCP server for Python AI course with get_info and get_greetings endpoints
Server has 5 tools with basic descriptions and parameter schemas, but significant quality gaps limit production readiness. All tools are READ_ONLY and well-intentioned, but descriptions lack LLM-optimization guidance, parameter descriptions are minimal, and output schemas are undocumented. No error recovery guidance, no input validation hints, and no tool annotations. Naming is clear (verb_noun pattern) but descriptions average ~120 chars, below the 194-char baseline for A+ tools. Parameter descriptions exist but are sparse (e.g., 'Path to the workspace folder' lacks format/constraint guidance). No documented output structure for complex tools like get_workspace_info and get_repo_info.
Get the diff between the current branch and a target branch. Args: repo_path: Path to the Git repository (can use ${workspaceFolder} in VS Code) target_branch: The target branch to compare against (default: main) Returns the diff between the current branch and the target branch.
Get a personalized greeting. Args: name: The name of the person to greet Returns a friendly greeting that includes the provided name.
Get information about this MCP server. Returns information about the purpose and capabilities of this demo MCP server.
Get basic information about a Git repository status. Args: repo_path: Path to the Git repository (can use ${workspaceFolder} in VS Code) Returns basic Git repository status information.
Get information about a workspace folder. Args: workspace_folder: Path to the workspace folder (can use ${workspaceFolder} in VS Code) Returns information about the files in the workspace, including counts and statistics.
Output schemas undocumented for complex tools. get_workspace_info and get_repo_info return free-text strings with no structured field definitions. LLMs cannot parse results reliably or chain to downstream tools.
Parameter descriptions lack format/constraint guidance. 'Path to the workspace folder' does not specify: absolute vs relative, required format, or error handling for invalid paths. LLMs cannot validate inputs before calling.
No error recovery guidance in descriptions. Tools return error strings (e.g., 'Error: The path does not exist') but descriptions do not tell LLMs what to do next or which tools to call for recovery.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 49 | <=2025-11-25 | v2 |
Tool descriptions are generic and lack LLM-optimization. Average length ~120 chars vs 194-char baseline. Descriptions do not explain WHEN to use each tool vs alternatives or what downstream actions are enabled by the result.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). All tools are READ_ONLY but this is not declared in schema, forcing LLMs to infer safety from descriptions alone.