Skip to main content
Glama
squirrelogic

Feature-Discussion MCP Server

by squirrelogic

begin_feature_discussion

Start a new feature discussion to explore implementation options, architecture considerations, and best practices with AI guidance.

Instructions

Start a new feature discussion

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
titleYesTitle or name of the feature

Implementation Reference

  • The switch case handler that executes the 'begin_feature_discussion' tool. It generates a feature ID, initializes the feature discussion object with the provided title and sets the first prompt, creates discussion context, and returns a response with the feature ID and initial prompt.
    case "begin_feature_discussion": {
      const { title } = request.params.arguments as any;
      const id = `f${Object.keys(featureDiscussions).length + 1}`;
    
      // Initialize new feature discussion
      featureDiscussions[id] = {
        id,
        title,
        description: "",
        requirements: [],
        status: 'in-discussion',
        createdAt: new Date().toISOString(),
        updatedAt: new Date().toISOString(),
        currentPrompt: FEATURE_DISCUSSION_PROMPTS[0].id
      };
    
      // Initialize discussion context
      discussionContexts[id] = {
        previousDecisions: [],
        relatedFeatures: [],
        technicalConstraints: [],
        conversationHistory: []
      };
    
      // Return the first prompt
      return {
        content: [
          {
            type: "text",
            text: `Feature discussion started for: ${title}\n\nFeature ID: ${id}\n\nFirst question:\n${FEATURE_DISCUSSION_PROMPTS[0].message}`
          }
        ]
      };
    }
  • src/index.ts:175-188 (registration)
    Registers the 'begin_feature_discussion' tool by defining its name, description, and input schema (requiring a 'title' string) in the ListToolsRequestSchema handler.
    {
      name: "begin_feature_discussion",
      description: "Start a new feature discussion",
      inputSchema: {
        type: "object",
        properties: {
          title: {
            type: "string",
            description: "Title or name of the feature"
          }
        },
        required: ["title"]
      }
    },
  • Defines the array of sequential prompts used throughout feature discussions. The 'begin_feature_discussion' tool initializes new discussions with the first prompt from this array ('initial_description').
    const FEATURE_DISCUSSION_PROMPTS: FeaturePrompt[] = [
      {
        id: "initial_description",
        message: "Please provide a brief description of the feature you'd like to discuss.",
        field: "description"
      },
      {
        id: "business_value",
        message: "What business value does this feature provide? How does it benefit users or stakeholders?",
        field: "businessValue"
      },
      {
        id: "target_users",
        message: "Who are the target users for this feature?",
        field: "targetUsers"
      },
      {
        id: "requirements",
        message: "What are the key requirements or constraints for this feature?",
        field: "requirements"
      },
      {
        id: "success_criteria",
        message: "What are the success criteria for this feature? How will we know it's working as intended?",
        field: "successCriteria"
      },
      {
        id: "technical_approach",
        message: "Do you have any specific technical approach in mind for implementing this feature?",
        field: "technicalApproach"
      },
      {
        id: "risks",
        message: "Are there any potential risks or challenges we should consider?",
        field: "risks"
      },
      {
        id: "timeline",
        message: "What's the desired timeline or priority for this feature?",
        field: "timeline"
      }
    ];

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/squirrelogic/mcp-feature-discussion'

If you have feedback or need assistance with the MCP directory API, please join our Discord server