JSONL Pipeline Validator

Find malformed JSON lines, duplicate object keys, missing required top-level fields and duplicate string IDs before the next agent pipeline step. No repair or external calls.

Input and output

At most 1 MB of UTF-8 text and 5,000 physical lines. Syntax diagnostics plus optional top-level key presence and string-ID uniqueness. Duplicate object keys are flagged. No repair, full JSON Schema validation, type validation or destination API acceptance guarantee. Values and IDs are omitted from output.

Fictional input example:

{
  "jsonlText": "{\"id\":\"one\",\"task\":\"example\"}\n{\"id\":\"one\"}\ninvalid\n",
  "requiredKeys": [
    "id",
    "task"
  ],
  "idKey": "id"
}

Fictional report example (the dataset returns an array containing this report):

{
  "schemaVersion": "1.0",
  "tool": "jsonl-validator",
  "inputSha256": "e267419d6f00e3d37458f0d9522156c856556f0c8754080c00102b10855dbb79",
  "inputBytes": 51,
  "summary": {
    "lines": 3,
    "validLines": 0,
    "reviewLines": 3
  },
  "duplicateIdLineGroups": [
    [
      1,
      2
    ]
  ],
  "results": [
    {
      "line": 1,
      "status": "needs_review",
      "reasons": [
        "duplicate_id"
      ],
      "duplicateKeyCount": 0
    },
    {
      "line": 2,
      "status": "needs_review",
      "reasons": [
        "missing_required_keys",
        "duplicate_id"
      ],
      "duplicateKeyCount": 0
    },
    {
      "line": 3,
      "status": "needs_review",
      "reasons": [
        "invalid_json"
      ],
      "duplicateKeyCount": 0
    }
  ],
  "boundaries": "Diagnostics only; no repairs, source fetching, schema/type validation, or downstream API acceptance guarantee. Duplicate JSON object keys are flagged as ambiguous. Values and IDs are omitted from the report."
}

The output is one report item in the default dataset, not one billed item per source row. No external source is fetched. Use authorized nonsensitive or redacted input. Input and results are stored by Apify under its retention/access rules; this is not a browser-local service.

Calling from an agent

After publication, discover this Actor through Apify Store / the Apify MCP server, inspect its input schema and current pricing, then run it with a bounded budget. Retrieve the default dataset only after a successful run. Check review flags; a completed diagnostic report may describe invalid input records. Never assume reviewed data is safe to import or execute.

Pricing and delivery

Trial price: USD 0.05 per completed report, not per input row. Check the current Store pricing before running; that listing is authoritative. One paid event report, with no paid startup/default-dataset events and no platform-usage passthrough. Set a spending budget before starting. Demand and margin remain unvalidated. The adapter supports FREE/unpriced testing and this single-event PPE model and rejects other paid configurations.

Invalid batch input fails before our custom charge. Reports containing row-level review findings are completed reports. Storage and billing are not transactional: a delivery/billing error is marked uncertain where possible and is never retried automatically. Existing dataset/status evidence blocks restarting the same run. Inspect delivery and charges before any manual retry; a new run is a new purchase.

Built by SuperFresh Labs. No customer results, usage or income claimed.