continue_batch

Instructions

Implementation Reference

• src/sqli_mcp/server.py:664-713 (handler)

  The continue_batch async function that handles continuing a batch scan. It retrieves the pending scan state from the pending_scans dict by batch_id, builds the remaining URLs, calls scan_urls_batch again with the stored parameters, and cleans up when no more URLs remain.

  async def continue_batch(batch_id: str) -> dict:
      """
      Continue scanning remaining URLs from a previous batch.
      Use this when scan_urls_batch returns has_more=True.
      
      Args:
          batch_id: Batch ID from a previous scan_urls_batch call
      
      Returns:
          Next batch of scan results
      """
      if batch_id not in pending_scans:
          return {"error": f"No pending scans for batch {batch_id}. Batch may be complete or expired."}
      
      pending = pending_scans[batch_id]
      remaining = pending["remaining_urls"]
      
      if not remaining:
          del pending_scans[batch_id]
          return {"message": "All URLs in this batch have been scanned", "batch_id": batch_id}
      
      # Build URL string for the next batch
      urls_str = "\n".join(remaining)
      
      # Scan next batch
      result = await scan_urls_batch(
          urls=urls_str,
          method=pending["method"],
          injection_types=pending["injection_types"],
          database_types=pending["database_types"],
          headers=pending["headers"],
          cookies=pending["cookies"],
          bearer_token=pending["bearer_token"],
          proxy_url=pending["proxy_url"],
          verify_ssl=pending["verify_ssl"],
          waf_bypass=pending["waf_bypass"],
          concurrency=pending["concurrency"],
          timeout=pending["timeout"],
          quick_mode=pending["quick_mode"],
          max_urls_per_batch=pending["max_urls_per_batch"]
      )
      
      # Update the original batch_id reference
      result["original_batch_id"] = batch_id
      
      # Clean up if complete
      if not result.get("has_more", False) and batch_id in pending_scans:
          del pending_scans[batch_id]
      
      return result

• src/sqli_mcp/server.py:663-664 (registration)

  The @mcp.tool() decorator that registers continue_batch as an MCP tool on line 663, just before the function definition.

  @mcp.tool()
  async def continue_batch(batch_id: str) -> dict:

• src/sqli_mcp/server.py:485-485 (helper)

  The pending_scans global dict that stores pending scan state (remaining URLs and parameters) for continuation by continue_batch.

  pending_scans: dict[str, dict] = {}

• src/sqli_mcp/server.py:580-597 (helper)

  The code in scan_urls_batch that populates pending_scans with the remaining URLs and all parameters so continue_batch can pick them up later.

  if remaining_urls:
      pending_scans[batch_id] = {
          "remaining_urls": remaining_urls,
          "method": method,
          "injection_types": injection_types,
          "database_types": database_types,
          "headers": headers,
          "cookies": cookies,
          "bearer_token": bearer_token,
          "proxy_url": proxy_url,
          "verify_ssl": verify_ssl,
          "waf_bypass": waf_bypass,
          "concurrency": concurrency,
          "timeout": timeout,
          "quick_mode": quick_mode,
          "max_urls_per_batch": max_urls_per_batch,
          "all_results": []
      }