JSON Text Health Diagnoser with Parse-Error Quarantine

How it works

1. 1.A Code node emits seven fictional {id, raw} records.
2. 2.The parse node checks raw type, then runs JSON.parse in try/catch per record.
3. 3.IF routes on parse_ok.
4. 4.Two Set nodes add status valid or quarantined, keeping other fields.
5. 5.Merge appends both branches; a final Code node builds one report sorted by index.

Use cases

• Checking LLM-generated JSON before downstream use.
• Quarantining hand-copied payloads with unescaped quotes.

Requirements

• No credentials needed.
• Adjust RAW_FIELD, ID_FIELD, PREVIEW_MAX and ALLOWED_ROOT_TYPES in the parse node.

Setup and testing in n8n

1. Import the JSON into n8n.
2. Click Execute Workflow.
3. Confirm 3 valid and 4 quarantined records in the report.
4. Add a record without id or raw and confirm missing-id or non_string_input classification.
5. Set ALLOWED_ROOT_TYPES to ['object'] and confirm the array and number get quarantined.

Service Value

Ready-made workflow template for automation delivery and service execution.