quality.getJobStatus
Check status and retrieve results for async quality evaluation jobs submitted via batch processing. Monitor progress percentage, item counts, and access result summaries or error details.
Instructions
Check the status of an async batch quality evaluation job.
Use this tool to poll for the status and results of a job that was submitted with quality.batch_evaluate.
Returns:
Job state (waiting, active, completed, failed)
Progress percentage (0-100)
Item counts (total, processed, success, failed)
Result summary when completed
Error details when failed
Note: When Redis is unavailable, falls back to LRU store (in-memory). Jobs in LRU store are not persisted across server restarts.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| job_id | Yes | The job ID returned by quality.batch_evaluate |
Implementation Reference
- The handler function 'qualityGetJobStatusHandler' implements the logic for 'quality.getJobStatus'. It attempts to retrieve job status from BullMQ (Redis) and falls back to an in-memory LRU store if the job is not found or Redis is unavailable.
export async function qualityGetJobStatusHandler( input: unknown ): Promise<QualityGetJobStatusOutput> { // router.tsから注入された_request_idを使用、フォールバックとして自動生成 const requestId = ((input as Record<string, unknown> | null)?._request_id as string | undefined) ?? generateRequestId(); if (isDevelopment()) { logger.info("[MCP Tool] quality.getJobStatus called", { hasInput: input !== null && input !== undefined, requestId, }); } // Validate input let validated: QualityGetJobStatusInput; try { validated = qualityGetJobStatusInputSchema.parse(input); } catch (error) { logger.warn("[MCP Tool] quality.getJobStatus validation error", { error: error instanceof Error ? error.message : String(error), requestId, }); return createErrorResponseWithRequestId( GET_QUALITY_JOB_STATUS_ERROR_CODES.VALIDATION_ERROR, error instanceof Error ? error.message : "Invalid input", requestId ); } const jobId = validated.job_id; // Check Redis availability const redisAvailable = await isRedisAvailable(); if (redisAvailable) { // Try BullMQ queue first const queue = createBatchQualityQueue(); try { const status = await getBatchQualityJobStatus(queue, jobId); if (status) { // Job found in BullMQ if (isDevelopment()) { logger.debug("[MCP Tool] quality.getJobStatus job found in BullMQ", { jobId, state: status.state, progress: status.progress, requestId, }); } const data: QualityGetJobStatusData = { jobId: status.jobId, status: mapBullMQStateToQualityState(status.state), progress: status.progress, totalItems: status.totalItems, processedItems: status.processedItems, successItems: status.successItems, failedItems: status.failedItems, timestamps: status.timestamps, }; // Add result if completed if (status.state === "completed" && status.result) { data.result = { jobId: status.result.jobId, success: status.result.success, totalItems: status.result.totalItems, processedItems: status.result.processedItems, successItems: status.result.successItems, failedItems: status.result.failedItems, results: status.result.results.map((r) => ({ index: r.index, success: r.success, data: r.data, error: r.error, })), processingTimeMs: status.result.processingTimeMs, completedAt: status.result.completedAt, error: status.result.error, }; } // Add error if failed if (status.state === "failed" && status.error) { data.failedReason = status.error; } return { success: true, data, metadata: { request_id: requestId, redis_used: true, lru_fallback: false, }, }; } // Job not found in BullMQ, try LRU store as fallback if (isDevelopment()) { logger.debug("[MCP Tool] quality.getJobStatus job not found in BullMQ, trying LRU store", { jobId, requestId, }); } } finally { // Close queue connection await closeBatchQualityQueue(queue); } } // Try LRU store (either Redis unavailable or job not found in BullMQ) const lruJob = getBatchJob(jobId); if (lruJob) { // Job found in LRU store if (isDevelopment()) { logger.debug("[MCP Tool] quality.getJobStatus job found in LRU store", { jobId, status: lruJob.status, progress: lruJob.progress_percent, requestId, }); } const data: QualityGetJobStatusData = { jobId: lruJob.job_id, status: mapLRUStatusToQualityState(lruJob.status), progress: lruJob.progress_percent, totalItems: lruJob.total_items, processedItems: lruJob.processed_items, successItems: lruJob.success_items, failedItems: lruJob.failed_items, timestamps: { created: new Date(lruJob.created_at).getTime(), completed: lruJob.completed_at ? new Date(lruJob.completed_at).getTime() : undefined, }, }; // Add result if completed if (lruJob.status === "completed" && lruJob.results) { data.result = { jobId: lruJob.job_id, success: lruJob.failed_items === 0, totalItems: lruJob.total_items, processedItems: lruJob.processed_items, successItems: lruJob.success_items, failedItems: lruJob.failed_items, results: lruJob.results.map((r, index) => ({ index, success: true, data: r, })), completedAt: lruJob.completed_at, }; // Add errors if any if (lruJob.errors && lruJob.errors.length > 0) { data.result.results = [ ...(data.result.results ?? []), ...lruJob.errors.map((e) => ({ index: e.index, success: false, error: e.error, })), ].sort((a, b) => a.index - b.index); } } // Add error if failed if (lruJob.status === "failed" && lruJob.errors && lruJob.errors.length > 0) { data.failedReason = lruJob.errors[0]?.error?.message ?? "Unknown error"; } return { success: true, data, metadata: { request_id: requestId, redis_used: false, lru_fallback: true, }, }; } // Job not found in either BullMQ or LRU store if (isDevelopment()) { logger.debug("[MCP Tool] quality.getJobStatus job not found", { jobId, requestId }); } return createErrorResponseWithRequestId( GET_QUALITY_JOB_STATUS_ERROR_CODES.JOB_NOT_FOUND, `Job with ID ${jobId} not found. It may have expired (jobs are retained for 24 hours) or never existed.`, requestId ); } - Tool definition and input schema for 'quality.getJobStatus'.
export const qualityGetJobStatusToolDefinition = { name: "quality.getJobStatus", description: `Check the status of an async batch quality evaluation job. Use this tool to poll for the status and results of a job that was submitted with quality.batch_evaluate. Returns: - Job state (waiting, active, completed, failed) - Progress percentage (0-100) - Item counts (total, processed, success, failed) - Result summary when completed - Error details when failed Note: When Redis is unavailable, falls back to LRU store (in-memory). Jobs in LRU store are not persisted across server restarts.`, inputSchema: { type: "object" as const, properties: { job_id: { type: "string", format: "uuid", description: "The job ID returned by quality.batch_evaluate", }, }, required: ["job_id"], additionalProperties: false, }, annotations: { title: "Quality Get Job Status", readOnlyHint: true, idempotentHint: true, openWorldHint: false, }, }; - apps/mcp-server/src/tools/index.ts:614-615 (registration)Registration of the 'quality.getJobStatus' tool within the MCP server's tool registry.
// quality.getJobStatus（Phase 3-5 一括評価ジョブステータス確認） "quality.getJobStatus": qualityGetJobStatusHandler,