respond to comments

Author: smw-ms

eng/common/scripts/Invoke-ToolDescriptionEvaluator.ps1

@@ -1,47 +1,50 @@
 <#
 .SYNOPSIS
-Evaluates tool and prompt descriptions and produces a Markdown results report using Tool Description Evaluator from Mcp repo.
+    Evaluates tool and prompt descriptions and produces a Markdown results report using Tool Description Evaluator from Mcp repo.
 
 .DESCRIPTION
-This script builds and runs the ToolDescriptionEvaluator (.NET) against a set of tool and prompt definitions.
-It restores and compiles the evaluator, executes it with the provided JSON inputs, and emits a `results.md`
-report. The script supports configuration via parameters and environment variables for the embedding model
-service used during evaluation.
+    This script builds and runs the ToolDescriptionEvaluator (.NET) against a set of tool and prompt definitions.
+    It restores and compiles the evaluator, executes it with the provided JSON inputs, and emits a `results.md`
+    report. The script supports configuration via parameters and environment variables for the embedding model
+    service used during evaluation.
+
+.LINK
+    https://github.com/microsoft/mcp/tree/main/eng/tools/ToolDescriptionEvaluator
 
 .PARAMETER EvaluatorPath
-The path to the evaluator project root (or its `src` directory) that will be restored, built, and executed.
+    The path to the evaluator project root (or its `src` directory) that will be restored, built, and executed.
 
 .PARAMETER ToolsFilePath
-The path to the JSON file containing tool definitions to be evaluated.
+    The path to the JSON file containing tool definitions to be evaluated.
 
 .PARAMETER PromptsFilePath
-The path to the JSON file containing prompt definitions to be evaluated.
+    The path to the JSON file containing prompt definitions to be evaluated.
 
 .PARAMETER OutputFilePath
-The target folder where the generated `results.md` will be moved after the evaluator runs.
+    The target file path where the generated `results.md` will be moved after the evaluator runs.
 
 .PARAMETER AoaiEndpoint
-The full endpoint URL for the embedding model (e.g., Azure OpenAI embeddings) used by the evaluator.
+    The full endpoint URL for the embedding model (e.g., Azure OpenAI embeddings) used by the evaluator.
 
 .PARAMETER TextEmbeddingApiKey
-The API key used to authenticate with the embedding endpoint. Prefer providing this via a secure
-secret store or environment variable rather than hard-coding.
+    The API key used to authenticate with the embedding endpoint. Prefer providing this via a secure
+    secret store or environment variable rather than hard-coding.
 
 .NOTES
-- The evaluator emits `results.md` in the evaluator folder; this script moves it to `OutputFilePath`.
-- Requires .NET SDK available on PATH.
-- Set-StrictMode is enabled.
+    - The evaluator emits `results.md` in the evaluator folder; this script moves it to `OutputFilePath`.
+    - Requires .NET SDK available on PATH.
+    - Set-StrictMode is enabled.
 
 .EXAMPLE
-.\Run-Evaluator.ps1 `
-  -EvaluatorPath "C:\work\mcp\eng\tools\ToolDescriptionEvaluator\src" `
-  -ToolsFilePath "C:\work\azure-sdk-tools\tools\azsdk-cli\azure-sdk-tools.json" `
-  -PromptsFilePath "C:\work\azure-sdk-tools\tools\azsdk-cli\azure-sdk-prompts.json" `
-  -OutputFilePath "C:\work\azure-sdk-tools\tools\azsdk-cli" `
-  -AoaiEndpoint "https://<your-endpoint>/openai/deployments/text-embedding-3-large/embeddings?api-version=2023-05-15" `
-  -TextEmbeddingApiKey (Get-Secret -Name 'TextEmbeddingApiKey')
-
-Runs the evaluator with the specified tools and prompts files, then moves the generated results to the output path.
+    .\Invoke-ToolDescriptionEvaluator.ps1 `
+    -EvaluatorPath "C:\work\mcp\eng\tools\ToolDescriptionEvaluator\src" `
+    -ToolsFilePath "C:\work\azure-sdk-tools\tools\azsdk-cli\azure-sdk-tools.json" `
+    -PromptsFilePath "C:\work\azure-sdk-tools\tools\azsdk-cli\azure-sdk-prompts.json" `
+    -OutputFilePath "C:\work\azure-sdk-tools\tools\azsdk-cli" `
+    -AoaiEndpoint "https://<your-endpoint>/openai/deployments/text-embedding-3-large/embeddings?api-version=2023-05-15" `
+    -TextEmbeddingApiKey (Get-Secret -Name 'TextEmbeddingApiKey')
+
+    Runs the evaluator with the specified tools and prompts files, then moves the generated results to the output path.
 #>
 param (
     [Parameter(Mandatory = $true)]
@@ -66,24 +69,34 @@ param (
 
 Set-StrictMode -Version 3
 
+# Validate input paths
+$pathsToCheck = @{
+    "EvaluatorPath"  = $EvaluatorPath
+    "ToolsFilePath"  = $ToolsFilePath
+    "PromptsFilePath" = $PromptsFilePath
+    "OutputFilePath" = $OutputFilePath
+}
+
+foreach ($p in $pathsToCheck.GetEnumerator()) {
+    if (-not (Test-Path -Path $p.Value)) {
+        throw "Path does not exist for parameter '$($p.Key)': $($p.Value)"
+    }
+}
+
 # Build & run the evaluator
 Write-Host "Changing directory to evaluator: $EvaluatorPath"
 Push-Location $EvaluatorPath
 try { 
     $env:AOAI_ENDPOINT = $AoaiEndpoint
     $env:TEXT_EMBEDDING_API_KEY = $TextEmbeddingApiKey
 
-    Write-Host "Restoring packages..."
-    dotnet restore
-
-    Write-Host "Building (Release)..."
-    dotnet build --configuration Release
-
     Write-Host "Running Tool..."
-    dotnet run -- --tools-file "$ToolsFilePath" --prompts-file "$PromptsFilePath"
+    dotnet run --configuration Release -- --tools-file "$ToolsFilePath" --prompts-file "$PromptsFilePath"
 }
 finally {
     Pop-Location
+    Remove-Item Env:\AOAI_ENDPOINT -ErrorAction SilentlyContinue
+    Remove-Item Env:\TEXT_EMBEDDING_API_KEY -ErrorAction SilentlyContinue
 }
 
 # The tool emits results.md in the evaluator folder