check_job_status¶

Checks the progress of a scheduled batch calculation job. Returns counts only (completed, failed, pending) - no wallet data. Use this after predictive_fraud_batch or predictive_behaviour_batch to monitor progress before retrieving results.

Only call get_job_results when status is "completed" or "partial". If status is "processing" or "pending", inform the user and wait before calling again.

Both job_id and signature from the batch scheduling call are required. Never call this tool without both values.

MCP Endpoint: https://prediction.mcp.chainaware.ai/sse

Input Schema¶

Field	Type	Required	Description
`job_id`	string	Yes	The `job_id` returned when scheduling the batch job
`signature`	string	Yes	The `signature` returned when scheduling the batch job

Output Schema¶

{
  "job_id": "0fc5897a-ad64-4f21-88b5-1274d1cfec46",
  "status": "partial",
  "chain": "ETH",
  "total_items": 5,
  "completed_items": 1,
  "failed_items": 4,
  "pending_items": 0,
  "expires_at": "2026-07-01T13:51:20.000Z"
}

Field	Type	Description
`job_id`	string	The job identifier
`status`	string	`"pending"`, `"processing"`, `"partial"`, or `"completed"`
`chain`	string	Blockchain network the job was submitted for
`total_items`	integer	Total wallets submitted in the batch
`completed_items`	integer	Wallets successfully processed so far
`failed_items`	integer	Wallets that failed processing
`pending_items`	integer	Wallets not yet processed
`expires_at`	datetime	When the job and its results expire

Status values¶

Status	Meaning	Next action
`pending`	Job is queued, not yet started	Wait and check again later
`processing`	Job is actively running	Wait and check again later
`partial`	Some wallets completed, some still processing or failed	Can call `get_job_results` for completed items
`completed`	All wallets processed	Call `get_job_results` to retrieve all results

Code Examples¶

Node.js¶

const status = await client.callTool({
  name: "check_job_status",
  arguments: {
    job_id: "0fc5897a-ad64-4f21-88b5-1274d1cfec46",
    signature: "260866090d88bf61bdfb54f0533fe876bfd8ded7339691c50ada9de59a48124a"
  }
});

if (status.status === "completed" || status.status === "partial") {
  // Safe to call get_job_results
} else {
  console.log(`Job still running: ${status.completed_items}/${status.total_items} done`);
}

Python¶

status = await session.call_tool("check_job_status", {
    "job_id": "0fc5897a-ad64-4f21-88b5-1274d1cfec46",
    "signature": "260866090d88bf61bdfb54f0533fe876bfd8ded7339691c50ada9de59a48124a"
})

if status["status"] in ("completed", "partial"):
    # Safe to call get_job_results
    pass

Example Agent Prompts¶

"What is the status of job 0fc5897a-ad64-4f21-88b5-1274d1cfec46?"
"Is my batch fraud job done? Job ID: 0fc5897a... Signature: 260866..."
"How many wallets have been processed in my batch job?"

Error Codes¶

Code	Meaning
`403`	Invalid `apiKey` or `signature`
`400`	Malformed `job_id`
`500`	Temporary backend failure - retry after a short delay

predictive_fraud_batch - schedule a batch fraud detection job
predictive_behaviour_batch - schedule a batch behavioural profiling job
get_job_results - retrieve results once status is completed or partial

See also: Prediction MCP Overview | Setup Guide