Add support for MCP prompts #336
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Important Review skippedReview was skipped due to path filters ⛔ Files ignored due to path filters (19)
CodeRabbit blocks several paths by default. You can override this behavior by explicitly including those paths in the path filters. For example, including You can disable this status message by setting the ✨ Finishing Touches🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
|
it looks good and reuse the pattern from tools, so it's as easy and same way of using it. |
Ellerbach
reviewed
Jul 23, 2025
josesimoes
force-pushed
the
add-mcp-prompts
branch
from
72469eb to
59c00cd
Compare
Ellerbach
reviewed
Jul 23, 2025
| { | ||
| PromptMetadata promptMetadata = (PromptMetadata)promtps[promptName]; | ||
| MethodInfo method = promptMetadata.Method; | ||
| Console.WriteLine($"Prompt name: {promptName}, method: {method.Name}"); |
| public static string AnotherSimplePrompt() | ||
| { | ||
| return "This is another prompt example"; | ||
| } |
josesimoes
force-pushed
the
add-mcp-prompts
branch
4 times, most recently
from
5355efd to
cbc1db4
Compare
- Add prompt registry class and methods. - Add prompt message, prompt parameter and roles classes. - Extract common methods to RegistryBase class. - Adjust McpToolRegistry accordingly. - Add tests for simple prompts. - Update MCP client test to list available prompts.
josesimoes
force-pushed
the
add-mcp-prompts
branch
from
cbc1db4 to
e64bc4e
Compare
Ellerbach
reviewed
Jul 28, 2025
| /// <inheritdoc/> | ||
| public override string ToString() | ||
| { | ||
| return $"{{\"name\":\"{Name}\",\"description\":\"{Description}\",\"required\":\"true\"}}"; |
There was a problem hiding this comment.
in the case you have multiple prompts in the same function, you've set them all to be required. Not sure it can smoothly work in that case. Maybe only 1 can be mandatory or leaving the choice in the attribute (with true as default)
There was a problem hiding this comment.
As the parameter name is not available in reflection, there is no way to tell which parameter is which, meaning no clean way to identify the parameters. That's why I've added this "by design" rule: if a parameter is declared it has to be mandatory and included in the call.
Ellerbach
reviewed
Jul 30, 2025
| @@ -46,7 +46,14 @@ | |||
| // Load them as AI functions in the kernel | |||
| #pragma warning disable SKEXP0001 | |||
| kernel.Plugins.AddFromFunctions("nanoFramework", tools.Select(aiFunction => aiFunction.AsKernelFunction())); | |||
| // -- | |||
|
|
|||
| // Check available prompts | |||
There was a problem hiding this comment.
you can use this instead those lines:
//
// 4. Register prompts from MCP
//
try
{
var prompts = await mcpToolboxClient.ListPromptsAsync().ConfigureAwait(false);
Console.WriteLine("// Available prompts:");
foreach (var p in prompts)
{
Console.WriteLine($"{p.Name}: {p.Description}");
var promptResult = await mcpToolboxClient.GetPromptAsync(p.Name, new Dictionary<string, object>());
var promptTemplate = string.Join("\n", promptResult.Messages.Select(m => m.Content));
var semanticFunction = KernelFunctionFactory.CreateFromPrompt(
promptTemplate: promptTemplate,
executionSettings: (PromptExecutionSettings?)null, // Explicit cast to resolve ambiguity
functionName: p.Name,
description: p.Description,
templateFormat: "semantic-kernel"
);
kernel.Plugins.AddFromFunctions(p.Name, new[] { semanticFunction });
}
Console.WriteLine("// --");
}
catch (Exception ex)
{
Console.WriteLine($"Error loading prompts: {ex.Message}");
}| { | ||
| return new PromptMessage[] | ||
| { | ||
| new PromptMessage($"Please perform the following steps:\r\n1. Call GetDefaultPerson() -> person.\r\n2. Call GetDefaultAddress() -> address.\r\n3. If person.Age > {ageThreshold} then set label = \"senior\"; otherwise set label = \"junior\".\r\n4. Call ProcessPerson(person) -> processed.\r\n5. Return a JSON object:\r\n{{ \"name\": person.Name, \"age\": person.Age, \"label\": label, \"address\": address, \"processed\": processed }}") |
There was a problem hiding this comment.
everything needs to be string encoded. so, this will break. This will actually work (or manual encoding):
Suggested change
| new PromptMessage($"Please perform the following steps:\r\n1. Call GetDefaultPerson() -> person.\r\n2. Call GetDefaultAddress() -> address.\r\n3. If person.Age > {ageThreshold} then set label = \"senior\"; otherwise set label = \"junior\".\r\n4. Call ProcessPerson(person) -> processed.\r\n5. Return a JSON object:\r\n{{ \"name\": person.Name, \"age\": person.Age, \"label\": label, \"address\": address, \"processed\": processed }}") | |
| new PromptMessage(JsonSerializer.SerializeObject($"Please perform the following steps:\r\n1. Call GetDefaultPerson() → person.\r\n2. Call GetDefaultAddress() → address.\r\n3. If person.Age > {ageThreshold ?? string.Empty} then set label = \"senior\"; otherwise set label = \"junior\".\r\n4. Call ProcessPerson(person) → processed.\r\n5. Return a JSON object:\r\n{{ \"name\": person.Name, \"age\": person.Age, \"label\": label, \"address\": address, \"processed\": processed }}").TrimStart('"').TrimEnd('"')) |
There was a problem hiding this comment.
For performance, I would prefer a manual encoded string, so there is no need to call the Json serializer.
| arguments.Values.CopyTo(methodParams, 0); | ||
| } | ||
|
|
||
| PromptMessage[] result = (PromptMessage[])method.Invoke(null, methodParams); |
There was a problem hiding this comment.
As the LLM does not really necessarily respect the number of arguments, we can do something like this:
Suggested change
| PromptMessage[] result = (PromptMessage[])method.Invoke(null, methodParams); | |
| PromptMessage[] result; | |
| try | |
| { | |
| result = (PromptMessage[])method.Invoke(null, methodParams); | |
| } | |
| catch (ArgumentException) | |
| { | |
| // Trying with 1 parameter as LLM may have forgot it | |
| result = (PromptMessage[])method.Invoke(null, new object[] { null }); | |
| } |
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Motivation and Context
How Has This Been Tested?
Screenshots
Types of changes
Checklist: