Skip to main content
Glama
Hug0x0

mcp-reunion

reunion_search_admin_directory

Locate public service counters in La Réunion (mairies, CAF, écoles, etc.) by free-text, commune, or type. Returns contact details and address.

Instructions

Search the Annuaire de l'Administration: local counters of public services in La Réunion (town halls / mairies, CCAS, CAF, Pôle emploi, sub-préfectures, tax offices, schools, etc.). Returns name, type (pivotlocal), full address, phone, email, website, opening hours notes, INSEE code, EPCI. Source: Service-Public.fr / DILA via data.regionreunion.com.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
queryNoFree-text search across name, address, services
communeNoCommune name prefix match (e.g. "Saint-Denis")
pivot_localNoService type prefix match. Examples: "mairie", "ccas", "caf", "pole_emploi", "sous_prefecture", "tresorerie", "ecole", "college", "lycee"
limitNoMax counters to return (1-200, default 50)

Implementation Reference

  • The handler function for the 'reunion_search_admin_directory' tool. Takes optional query, commune, pivot_local, and limit parameters; queries the OpenDataSoft API dataset 'annuaire-de-ladministration-base-de-donnees-localespublic'; maps results to a structured response with id, insee_code, updated, pivot_local, name, address_line, postal_code, commune, email, url, opening_notes, and epci fields.
      async ({ query, commune, pivot_local, limit }) => {
        try {
          const data = await client.getRecords<RecordObject>(DATASET_ADMIN_DIR, {
            where: buildWhere([
              query ? `search(${quote(query)})` : undefined,
              commune ? `adresse_nomcommune LIKE ${quote(`${commune}%`)}` : undefined,
              pivot_local ? `pivotlocal LIKE ${quote(`${pivot_local}%`)}` : undefined,
            ]),
            limit,
          });
          return jsonResult({
            total_counters: data.total_count,
            counters: data.results.map((row) => ({
              id: pickString(row, ['id']),
              insee_code: pickString(row, ['codeinsee']),
              updated: pickString(row, ['datemiseajour']),
              pivot_local: pickString(row, ['pivotlocal']),
              name: pickString(row, ['nom']),
              address_line: pickString(row, ['adresse_ligne']),
              postal_code: pickString(row, ['adresse_codepostal']),
              commune: pickString(row, ['adresse_nomcommune']),
              email: pickString(row, ['coordonneesnum_email']),
              url: pickString(row, ['coordonneesnum_url']),
              opening_notes: pickString(row, ['ouverture_plagej_note']),
              epci: pickString(row, ['nom_epci']),
            })),
          });
        } catch (error) {
          return errorResult(error instanceof Error ? error.message : 'Failed to search admin directory');
        }
      }
    );
  • Zod schema definitions for the input parameters of the tool: query (optional string), commune (optional string), pivot_local (optional string), and limit (optional number, default 50, min 1, max 200).
    {
      query: z.string().optional().describe('Free-text search across name, address, services'),
      commune: z.string().optional().describe('Commune name prefix match (e.g. "Saint-Denis")'),
      pivot_local: z.string().optional().describe('Service type prefix match. Examples: "mairie", "ccas", "caf", "pole_emploi", "sous_prefecture", "tresorerie", "ecole", "college", "lycee"'),
      limit: z.number().int().min(1).max(200).default(50).describe('Max counters to return (1-200, default 50)'),
    },
  • Registration of the tool on the MCP server via server.tool('reunion_search_admin_directory', ...). The tool description explains it searches the Annuaire de l'Administration for local public service counters in La Réunion.
    server.tool(
      'reunion_search_admin_directory',
      'Search the Annuaire de l\'Administration: local counters of public services in La Réunion (town halls / mairies, CCAS, CAF, Pôle emploi, sub-préfectures, tax offices, schools, etc.). Returns name, type (pivotlocal), full address, phone, email, website, opening hours notes, INSEE code, EPCI. Source: Service-Public.fr / DILA via data.regionreunion.com.',
      {
        query: z.string().optional().describe('Free-text search across name, address, services'),
        commune: z.string().optional().describe('Commune name prefix match (e.g. "Saint-Denis")'),
        pivot_local: z.string().optional().describe('Service type prefix match. Examples: "mairie", "ccas", "caf", "pole_emploi", "sous_prefecture", "tresorerie", "ecole", "college", "lycee"'),
        limit: z.number().int().min(1).max(200).default(50).describe('Max counters to return (1-200, default 50)'),
      },
      async ({ query, commune, pivot_local, limit }) => {
        try {
          const data = await client.getRecords<RecordObject>(DATASET_ADMIN_DIR, {
            where: buildWhere([
              query ? `search(${quote(query)})` : undefined,
              commune ? `adresse_nomcommune LIKE ${quote(`${commune}%`)}` : undefined,
              pivot_local ? `pivotlocal LIKE ${quote(`${pivot_local}%`)}` : undefined,
            ]),
            limit,
          });
          return jsonResult({
            total_counters: data.total_count,
            counters: data.results.map((row) => ({
              id: pickString(row, ['id']),
              insee_code: pickString(row, ['codeinsee']),
              updated: pickString(row, ['datemiseajour']),
              pivot_local: pickString(row, ['pivotlocal']),
              name: pickString(row, ['nom']),
              address_line: pickString(row, ['adresse_ligne']),
              postal_code: pickString(row, ['adresse_codepostal']),
              commune: pickString(row, ['adresse_nomcommune']),
              email: pickString(row, ['coordonneesnum_email']),
              url: pickString(row, ['coordonneesnum_url']),
              opening_notes: pickString(row, ['ouverture_plagej_note']),
              epci: pickString(row, ['nom_epci']),
            })),
          });
        } catch (error) {
          return errorResult(error instanceof Error ? error.message : 'Failed to search admin directory');
        }
      }
    );
  • ReunionClient class with getRecords method used by the handler to fetch data from the OpenDataSoft API. The client handles caching, retries, and timeouts.
    export class ReunionClient {
      private readonly baseUrl = 'https://data.regionreunion.com/api/explore/v2.1/';
      private readonly timeout = 30000;
      private readonly maxRetries = 2;
      private readonly metadataCache = new Map<string, Promise<DatasetMetadata | undefined>>();
      private readonly recordsCache = new Map<string, { value: unknown; expiresAt: number }>();
    
      /**
       * Fetch records from a dataset
       */
      async getRecords<T extends RecordObject = RecordObject>(
        datasetId: string,
        params: ODSQueryParams = {}
      ): Promise<ODSResponse<T>> {
        const url = this.buildUrl(`/catalog/datasets/${datasetId}/records`, params);
    
        if (REFERENTIAL_DATASETS.has(datasetId)) {
          const now = Date.now();
          const cached = this.recordsCache.get(url);
          if (cached && cached.expiresAt > now) {
            return cached.value as ODSResponse<T>;
          }
          const value = await this.fetchJson<ODSResponse<T>>(url);
          this.recordsCache.set(url, { value, expiresAt: now + REFERENTIAL_TTL_MS });
          return value;
        }
    
        return this.fetchJson<ODSResponse<T>>(url);
      }
    
      /**
       * Clear the in-memory caches. Intended for tests.
       */
      clearCaches(): void {
        this.metadataCache.clear();
        this.recordsCache.clear();
      }
    
      /**
       * Fetch aggregated data from a dataset
       */
      async getAggregates<T extends RecordObject = RecordObject>(
        datasetId: string,
        select: string,
        options: {
          where?: string;
          groupBy?: string;
          orderBy?: string;
          limit?: number;
        } = {}
      ): Promise<ODSResponse<T>> {
        const params: Record<string, string | number | undefined> = { select };
        if (options.where) params.where = options.where;
        if (options.groupBy) params.group_by = options.groupBy;
        if (options.orderBy) params.order_by = options.orderBy;
        if (options.limit !== undefined) params.limit = options.limit;
    
        const url = this.buildUrl(`/catalog/datasets/${datasetId}/aggregates`, params);
        return this.fetchJson<ODSResponse<T>>(url);
      }
    
      /**
       * Search across all datasets
       */
      async searchDatasets(query: string): Promise<CatalogResponse> {
        const url = this.buildUrl('/catalog/datasets', {
          where: `search(${quote(query)})`,
          limit: 20,
        });
        return this.fetchJson<CatalogResponse>(url);
      }
    
      /**
       * List datasets with an optional raw ODSQL where clause.
       */
      async listDatasets(
        options: { where?: string; limit?: number; offset?: number } = {}
      ): Promise<CatalogResponse> {
        const url = this.buildUrl('/catalog/datasets', {
          where: options.where,
          limit: options.limit ?? 20,
          offset: options.offset,
        });
        return this.fetchJson<CatalogResponse>(url);
      }
    
      /**
       * Fetch dataset metadata from the catalog
       */
      async getDatasetMetadata(datasetId: string): Promise<DatasetMetadata | undefined> {
        if (!this.metadataCache.has(datasetId)) {
          const promise = this.fetchJson<CatalogResponse>(
            this.buildUrl('/catalog/datasets', {
              where: `dataset_id = ${quote(datasetId)}`,
              limit: 1,
            })
          ).then((data) => data.results[0]);
    
          this.metadataCache.set(datasetId, promise);
        }
    
        return this.metadataCache.get(datasetId);
      }
    
      /**
       * Check whether a dataset currently exists in the public catalog
       */
      async datasetExists(datasetId: string): Promise<boolean> {
        return Boolean(await this.getDatasetMetadata(datasetId));
      }
    
      /**
       * Resolve the first matching field name for a dataset
       */
      async resolveField(
        datasetId: string,
        candidates: string[]
      ): Promise<string | undefined> {
        const metadata = await this.getDatasetMetadata(datasetId);
        const fields = metadata?.fields ?? [];
    
        if (fields.length === 0) {
          return candidates[0];
        }
    
        const byNormalizedName = new Map(
          fields.map((field) => [normalizeText(field.name), field.name] as const)
        );
    
        for (const candidate of candidates) {
          const direct = byNormalizedName.get(normalizeText(candidate));
          if (direct) {
            return direct;
          }
        }
    
        const fieldNames = fields.map((field) => field.name);
        for (const candidate of candidates) {
          const normalizedCandidate = normalizeText(candidate);
          const partial = fieldNames.find((fieldName) =>
            normalizeText(fieldName).includes(normalizedCandidate)
          );
          if (partial) {
            return partial;
          }
        }
    
        return candidates[0];
      }
    
      /**
       * Build URL with query parameters
       */
      private buildUrl(
        path: string,
        params: Record<string, string | number | undefined>
      ): string {
        const normalizedPath = path.startsWith('/') ? path.slice(1) : path;
        const url = new URL(normalizedPath, this.baseUrl);
    
        for (const [key, value] of Object.entries(params)) {
          if (value !== undefined && value !== null && value !== '') {
            url.searchParams.set(key, String(value));
          }
        }
    
        return url.toString();
      }
    
      /**
       * Execute HTTP request with retries and timeout handling
       */
      private async fetchJson<T>(url: string, remainingRetries = this.maxRetries): Promise<T> {
        const controller = new AbortController();
        const timeoutId = setTimeout(() => controller.abort(), this.timeout);
    
        try {
          const response = await fetch(url, {
            method: 'GET',
            headers: {
              Accept: 'application/json',
              'User-Agent': 'mcp-reunion/1.0',
            },
            signal: controller.signal,
          });
    
          if (!response.ok) {
            const errorText = await response.text();
            if (response.status >= 500 && remainingRetries > 0) {
              await this.delay(250);
              return this.fetchJson<T>(url, remainingRetries - 1);
            }
            throw new Error(
              `API error ${response.status}: ${response.statusText}. ${errorText}`
            );
          }
    
          return (await response.json()) as T;
        } catch (error) {
          if (error instanceof Error) {
            if (error.name === 'AbortError') {
              throw new Error(`Request timeout after ${this.timeout}ms`);
            }
    
            if (remainingRetries > 0 && this.isRetryableError(error)) {
              await this.delay(250);
              return this.fetchJson<T>(url, remainingRetries - 1);
            }
    
            throw error;
          }
          throw new Error('Unknown error occurred');
        } finally {
          clearTimeout(timeoutId);
        }
      }
    
      private isRetryableError(error: Error): boolean {
        return /fetch failed|ECONNRESET|ETIMEDOUT|ENOTFOUND|EAI_AGAIN/i.test(error.message);
      }
    
      private async delay(ms: number): Promise<void> {
        await new Promise((resolve) => setTimeout(resolve, ms));
      }
    }
    
    // Singleton instance
    export const client = new ReunionClient();
  • Helper functions buildWhere (combines ODSQL conditions) and quote (escapes/literals strings) used by the handler to construct the WHERE clause for the API query.
    export function buildWhere(
      conditions: Array<string | undefined | null | false>
    ): string | undefined {
      const valid = conditions.filter((condition): condition is string => Boolean(condition));
      return valid.length > 0 ? valid.join(' AND ') : undefined;
    }
    
    /**
     * Escape a string literal for ODSQL
     */
    export function escapeOdSqlString(value: string): string {
      return value.replace(/'/g, "''");
    }
    
    /**
     * Quote an ODSQL string literal
     */
    export function quote(value: string): string {
      return `'${escapeOdSqlString(value)}'`;
    }
Behavior2/5

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

No annotations are provided, so the description carries the full burden. It discloses what the tool returns and the source, but does not mention side effects, authentication needs, rate limits, or explicitly state it is read-only.

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

Conciseness4/5

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

The description is two sentences and packs essential information without waste. It is front-loaded with the purpose, though a bit dense.

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

Completeness4/5

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

The description lists return fields (name, type, address, etc.) and source, compensating for the missing output schema. Parameter limit is explained. Minor missing details on ordering or pagination, but sufficient for a search tool.

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

Parameters3/5

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

Schema coverage is 100% and descriptions are adequate. The description adds some context like pivot_local examples and commune prefix match, but does not significantly enhance meaning beyond the schema.

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

Purpose5/5

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

The description clearly states the tool searches 'Annuaire de l'Administration' for local counters of public services in La Réunion, listing specific examples and return fields. It is distinct from siblings like reunion_search_schools or reunion_search_health_professionals.

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

Usage Guidelines3/5

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

The description implies usage for finding administrative offices and provides examples of service types, but does not explicitly state when to use this tool versus alternatives or provide when-not-to-use guidance.

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

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/Hug0x0/mcp-reunion'

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