Search Standards by Crosscutting Concept

Instructions

Implementation Reference

• src/server/index.ts:456-496 (handler)

  The main handler function for the search_by_crosscutting_concept tool. It takes 'concept' (a CCC value) and optional 'detail_level', fetches all standards from the database, filters by exact CCC match, formats the response, and returns JSON with token metadata.

    async ({ concept, detail_level }) => {
      try {
        ensureInitialized();
        const db = getDatabase();
        const allStandards = db.getAllStandards();
  
        // Filter by CCC name (exact match)
        const filtered = allStandards.filter(s => s.ccc.name === concept);
  
        const formattedStandards = formatResponseArray(filtered, detail_level as DetailLevel);
        const tokens = getTokenMetadata(concept, formattedStandards);
  
        const result = {
          concept,
          total: filtered.length,
          standards: formattedStandards,
          _metadata: { tokens }
        };
  
        return {
          content: [{
            type: 'text',
            text: JSON.stringify(result, null, 2)
          }]
        };
      } catch (error) {
        console.error('search_by_crosscutting_concept error:', error);
        return {
          content: [{
            type: 'text',
            text: JSON.stringify({
              error: 'Internal Error',
              message: error instanceof Error ? error.message : String(error),
              code: 'INTERNAL_ERROR'
            }, null, 2)
          }],
          isError: true
        };
      }
    }
  );

• src/server/index.ts:442-496 (registration)

  Registration of the 'search_by_crosscutting_concept' tool with the MCP server, including title, description, and input schema definition.

  server.registerTool(
    'search_by_crosscutting_concept',
    {
      title: 'Search Standards by Crosscutting Concept',
      description: 'Find all NGSS standards using a specific Crosscutting Concept (CCC). Examples: "Patterns", "Cause and Effect", "Systems and System Models", "Energy and Matter"',
      inputSchema: {
        concept: z.enum(CCC_VALUES)
          .describe('Crosscutting Concept name'),
        detail_level: z.enum(['minimal', 'summary', 'full'])
          .optional()
          .default('full')
          .describe('Response detail level: minimal (code, topic, PE 50 chars), summary (+ keywords top 3, PE 150 chars), full (complete standard)')
      }
    },
    async ({ concept, detail_level }) => {
      try {
        ensureInitialized();
        const db = getDatabase();
        const allStandards = db.getAllStandards();
  
        // Filter by CCC name (exact match)
        const filtered = allStandards.filter(s => s.ccc.name === concept);
  
        const formattedStandards = formatResponseArray(filtered, detail_level as DetailLevel);
        const tokens = getTokenMetadata(concept, formattedStandards);
  
        const result = {
          concept,
          total: filtered.length,
          standards: formattedStandards,
          _metadata: { tokens }
        };
  
        return {
          content: [{
            type: 'text',
            text: JSON.stringify(result, null, 2)
          }]
        };
      } catch (error) {
        console.error('search_by_crosscutting_concept error:', error);
        return {
          content: [{
            type: 'text',
            text: JSON.stringify({
              error: 'Internal Error',
              message: error instanceof Error ? error.message : String(error),
              code: 'INTERNAL_ERROR'
            }, null, 2)
          }],
          isError: true
        };
      }
    }
  );

• src/server/index.ts:447-454 (schema)

  Input schema for the tool: 'concept' is a required enum of CCC_VALUES, 'detail_level' is an optional enum ('minimal', 'summary', 'full') defaulting to 'full'.

  inputSchema: {
    concept: z.enum(CCC_VALUES)
      .describe('Crosscutting Concept name'),
    detail_level: z.enum(['minimal', 'summary', 'full'])
      .optional()
      .default('full')
      .describe('Response detail level: minimal (code, topic, PE 50 chars), summary (+ keywords top 3, PE 150 chars), full (complete standard)')
  }

• src/server/token-counter.ts:35-48 (helper)

  Helper function getTokenMetadata used to estimate input/output tokens for the response metadata.

  export function getTokenMetadata(input: string, output: unknown): {
    input_tokens: number;
    output_tokens: number;
    total_tokens: number;
  } {
    const inputTokens = estimateTokens(input);
    const outputTokens = estimateTokensForObject(output);
  
    return {
      input_tokens: inputTokens,
      output_tokens: outputTokens,
      total_tokens: inputTokens + outputTokens
    };
  }

• src/server/response-formatter.ts:83-88 (helper)

  Helper function formatResponseArray used to format standards based on the requested detail level (minimal, summary, or full).

  export function formatResponseArray(
    standards: Standard[],
    detailLevel: DetailLevel = 'full'
  ): (MinimalStandard | SummaryStandard | Standard)[] {
    return standards.map(standard => formatResponse(standard, detailLevel));
  }