Anonymize Dataset

Overview

When you send columns, rows, and an explicit per-column policy, we transform values deterministically (drop, null, constant, mask, truncate, hash, hmac_token, bucket, round, date_shift, generalize, shuffle) and return transformed rows plus a per-column report. We validate the whole policy before touching any row. Same data, policy, secret, and seed yield identical output — do not use this as proof of legal anonymization or GDPR compliance.

Ad-hoc masking is inconsistent. We apply the policy you declare the same way every time.

When to use it

Scope

Non-goals

• Does not certify legal anonymization
• Does not auto-detect PII
• Does not return raw secrets

Properties

Capabilities

Cost & billing

Flat rate per invocation

File handling

How we handle your data

Input schema

Output schema

API contract

curl -sS -X POST "$BASE/v1/execute" -H "Authorization: Bearer $KEY" -H 'Content-Type: application/json' -d '{"skillKey":"data.anonymize_dataset","input":{"columns":["age"],"rows":[{"age":37}],"policy":[{"column":"age","action":"round","step":10}]}}'

import { Squillify } from "@squillify/sdk";

const sq = new Squillify({ apiKey: process.env.SQUILLIFY_API_KEY });

const { output } = await sq.execute("data.anonymize_dataset", {"rows":[{"name":"Alice","age":37}],"policy":[{"column":"name","action":"drop"},{"action":"bucket","boundaries":[0,18,30,45,65,200],"column":"age"}],"columns":["name","age"]});

Errors

Other limits

Limitations

• Does not certify legal anonymization
• Does not auto-detect PII
• Does not return raw secrets
• Claiming a dataset is legally anonymous
• Reversible tokens without a secret

How to call

POST /v1/execute (requires authentication)

{
  "skillKey": "data.anonymize_dataset",
  "input": {
    "rows": [
      {
        "name": "Alice",
        "age": 37
      }
    ],
    "policy": [
      {
        "column": "name",
        "action": "drop"
      },
      {
        "action": "bucket",
        "boundaries": [
          0,
          18,
          30,
          45,
          65,
          200
        ],
        "column": "age"
      }
    ],
    "columns": [
      "name",
      "age"
    ]
  }
}

Examples

{
  "rows": [
    {
      "name": "Alice",
      "age": 37
    }
  ],
  "policy": [
    {
      "column": "name",
      "action": "drop"
    },
    {
      "action": "bucket",
      "boundaries": [
        0,
        18,
        30,
        45,
        65,
        200
      ],
      "column": "age"
    }
  ],
  "columns": [
    "name",
    "age"
  ]
}

{
  "policyDigest": "50d8326c5d473e9e",
  "_meta": {
    "policyVersion": "anonymize-1",
    "rowsTransformed": 1
  },
  "rows": [
    {
      "age": "[30,45)"
    }
  ],
  "columns": [
    "age"
  ],
  "warnings": [],
  "columnStats": {
    "age": {
      "transformed": 1
    }
  }
}

{
  "rows": [
    {
      "age": 37
    }
  ],
  "policy": [
    {
      "action": "round",
      "step": 10,
      "column": "age"
    }
  ],
  "columns": [
    "age"
  ]
}

{
  "policyDigest": "f20333391323d3d1",
  "_meta": {
    "policyVersion": "anonymize-1",
    "rowsTransformed": 1
  },
  "rows": [
    {
      "age": 40
    }
  ],
  "columns": [
    "age"
  ],
  "warnings": [],
  "columnStats": {
    "age": {
      "transformed": 1
    }
  }
}

Recommended use cases

• Drop or mask identifiers before sharing
• Bucket ages or dates deterministically
• Pseudonymize with HMAC tokens when you supply a secret

Avoid for

• Claiming a dataset is legally anonymous
• Reversible tokens without a secret

Related skills

Composable with

Safety & privacy

• Runs without calling any external service.
• Your payloads are never stored permanently.
• Telemetry captures metadata only — never your payload contents.
• We never train models on your payloads.
• No human reviews your payloads.

Tags