Skip to main content
Glama
unctad-ai

eRegulations MCP Server

by unctad-ai

searchProcedures

Search for administrative procedures using specific keywords on the eRegulations MCP Server. Simplify access to structured data for AI models and user queries.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
keywordYesThe keyword or phrase to search for procedures. This will be wrapped in a JSON object with a 'keyword' property when sent to the API.

Implementation Reference

  • Creates and returns the MCP tool handler for 'searchProcedures', including input schema, description, and the async handler function that performs the search using ERegulationsApi, filters results to procedures, formats output using SearchProceduresFormatter, and returns structured text content.
    export function createSearchProceduresHandler(
      api: ERegulationsApi
    ): ToolHandler {
      return {
        name: ToolName.SEARCH_PROCEDURES,
        description: `Search for procedures by keyword or phrase. The search uses OR logic between words in the keyword phrase. For best results, prefer using a single, specific keyword whenever possible.`,
        inputSchema: zodToJsonSchema(SearchProceduresSchema),
        inputSchemaDefinition: SearchProceduresSchema,
        handler: async (args) => {
          // Use the inferred type for args
          const { keyword } = args as SearchProceduresArgs;
          logger.log(`Handling searchProcedures with keyword: ${keyword}`);
    
          try {
            const results = await api.searchProcedures(keyword);
    
            // Filter results to only include actual procedures based on links
            const filteredProcedures = results.filter(
              (item) =>
                item.links &&
                Array.isArray(item.links) &&
                item.links.some((link) => link && link.rel === "procedure")
            );
    
            logger.log(
              `searchProcedures API returned ${results.length} items, filtered to ${filteredProcedures.length} procedures`
            );
    
            // Format the filtered results using the search procedures formatter
            const formattedResult = formatters.searchProcedures.format(
              filteredProcedures, // Pass filtered procedures
              keyword // Pass keyword for context
            );
    
            logger.log(
              `searchProcedures found ${filteredProcedures.length} results`
            );
    
            // Always return only text content
            return {
              content: [
                {
                  type: "text",
                  text: formattedResult.text,
                },
              ],
            };
          } catch (error) {
            logger.error(
              `Error searching procedures with keyword "${keyword}":`,
              error
            );
            const errorMessage =
              error instanceof Error ? error.message : String(error);
            return {
              content: [
                {
                  type: "text",
                  text: `Error searching for procedures: ${errorMessage}`,
                },
              ],
            };
          }
        },
      };
    }
  • Zod input schema for the searchProcedures tool, defining the required 'keyword' string parameter with description.
    export const SearchProceduresSchema = z.object({
      keyword: z
        .string()
        .describe(
          "The keyword or phrase to search for procedures. This will be wrapped in a JSON object with a 'keyword' property when sent to the API."
        ),
    });
  • Registers the searchProcedures handler by including createSearchProceduresHandler(api) in the array returned by createHandlers, which initializes all MCP tool handlers.
    export function createHandlers(api: ERegulationsApi): ToolHandler[] {
      return [
        createListProceduresHandler(api),
        createGetProcedureDetailsHandler(api),
        createGetProcedureStepHandler(api),
        createSearchProceduresHandler(api),
      ];
    }
  • Core API method in ERegulationsApi class that performs the HTTP POST request to /Objectives/Search endpoint with the keyword, handles response, and returns array of matching objectives/procedures.
    async searchProcedures(
      keyword: string
    ): Promise<ObjectiveWithDescriptionBaseModel[]> {
      if (!keyword || typeof keyword !== "string") {
        throw new Error("Search keyword is required");
      }
    
      return this.fetchData<ObjectiveWithDescriptionBaseModel[]>(async () => {
        logger.log(`Searching objectives with keyword "${keyword}"...`);
    
        // Get the base URL at execution time
        const baseUrl = this.getBaseUrl();
    
        try {
          // Use POST for the search endpoint as specified in the API docs
          const url = `${baseUrl}/Objectives/Search`;
    
          // Create a specific axios instance for this POST request
          // Wrap keyword in an object as per the new API format
          const requestBody = { keyword };
          const response = await this.axiosInstance.post<
            ObjectiveWithDescriptionBaseModel[]
          >(url, JSON.stringify(requestBody), {
            headers: {
              "Content-Type": "application/json",
              Accept: "application/json",
            },
            timeout: REQUEST_CONFIG.TIMEOUT,
          });
    
          if (!response || !response.data) {
            logger.warn(`No objectives found for search keyword "${keyword}"`);
            return [];
          }
    
          // API should return ObjectiveWithDescriptionBaseModel[] directly
          if (Array.isArray(response.data)) {
            logger.log(
              `Found ${response.data.length} objective search results for "${keyword}"`
            );
            // No need to map or add searchKeyword, return data directly
            return response.data;
          } else {
            // Log unexpected responses, but still try to return an empty array
            logger.warn(
              `Unexpected search API response type for objectives: ${typeof response.data}. Expected Array.`
            );
            return [];
          }
        } catch (error) {
          logger.error(
            `Error searching objectives with keyword "${keyword}":`,
            error
          );
          // Don't expose internal errors directly, return empty array
          return [];
        }
      });
    }
  • Formatter class used in the searchProcedures handler to convert raw API search results into LLM-friendly formatted text and essential data summary.
    export class SearchProceduresFormatter
      implements DataFormatter<ObjectiveData[], FormattedProcedureList>
    {
      // Renamed class
      /**
       * Format procedure search results for LLM consumption.
       * @param results The search result data to format (ObjectiveWithDescriptionBaseModel[]).
       * @param keyword The search keyword used.
       * @returns Formatted procedure search results text and essential data.
       */
      public format(
        results: ObjectiveData[],
        keyword?: string
      ): FormattedProcedureList {
        // Renamed variable
        if (!results || !Array.isArray(results) || results.length === 0) {
          return {
            text: `No procedures found matching "${keyword || "search term"}"`,
            data: [],
          };
        }
    
        const formattedText = this.formatText(results, keyword); // Pass new params
        const essentialData = this.extractEssentialData(results); // Pass results
    
        return {
          text: formattedText,
          data: essentialData,
        };
      }
    
      /**
       * Extract essential data from search results (objectives) for procedure list.
       * @param results The full search results (objectives).
       * @returns A simplified array with essential fields (id, name, description).
       */
      private extractEssentialData(results: ObjectiveData[]): any[] {
        return results.map((res) => ({
          id: res.id,
          name: res.name,
          ...(res.description ? { description: res.description } : {}),
        }));
      }
    
      /**
       * Format search results (objectives) as human-readable text representing procedures.
       * @param results The search data to format (objectives).
       * @param keyword The search keyword used.
       * @returns Formatted text optimized for LLM context.
       */
      private formatText(results: ObjectiveData[], keyword?: string): string {
        const searchTerm = keyword ? ` for "${keyword}"` : "";
        const resultCount = results.length;
    
        let header = `Found ${resultCount} procedure${
          resultCount !== 1 ? "s" : ""
        }${searchTerm}:\n\n`; // Use "procedure" in text
    
        const shownResults = results; // Always show all results now
    
        if (shownResults.length > 0) {
          shownResults.forEach((res, index) => {
            const id = res.id || "N/A";
            const name = res.name || "Unknown";
    
            let description = "";
            if (res.description) {
              description = `\n   ${res.description}`;
            }
    
            header += `${index + 1}. ${name} (ID:${id})${description}\n`;
          });
    
          header += `\n\nTo get details about a specific procedure, use the getProcedureDetails tool with the procedure ID.`; // Keep instruction
        } else {
          header += "No procedures found.";
        }
    
        return header;
      }
    }
Behavior1/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Tool has no description.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness1/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Tool has no description.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness1/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Tool has no description.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters1/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Tool has no description.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose1/5

Does the description clearly state what the tool does and how it differs from similar tools?

Tool has no description.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines1/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Tool has no description.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Related Tools

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/unctad-ai/eregulations-mcp-server'

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