# Scientific Dataset Records - Semantic Parity Export

> Scientific Dataset Records - Semantic Parity Export is a paid API for AI agents from www.mahastrategies.com, paid per call via x402, $0.039338/call, status unknown (last checked 2026-10-20).

Compares parsed typed record projections of scientific dataset records for semantic parity, preserving pins and scope while ignoring field order, for public or synthetic data classes.

## Facts

- Endpoint: POST https://www.mahastrategies.com/api/v1/intel/scientific-dataset-records/export-semantic-parity?utm_source=zero.xyz
- Price: $0.039338/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-20
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/scientific-dataset-records-semantic-parity-export-ec326ff2
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_bvldFR59ehB4HlG8L5MN1

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability scientific-dataset-records-semantic-parity-export-ec326ff2 -d '<json body>'
```

Example prompt: Compare these two sets of scientific dataset records — up to 10 entries in the before array and 10 in the after array — for semantic parity using the 'public' data class with nonpersonal metadata only, and tell me if the typed projections match ignoring field order.

## When to prefer this

Use this endpoint when you need to confirm that two versions of scientific dataset records are semantically equivalent under typed projection — particularly when field ordering may differ but logical content should be the same. Prefer this over byte-level or hash-based comparison when record serialization order is non-deterministic. Best suited for public or synthetic (non-personal) dataset records with up to 32 entries per side.

## Known failure modes

- Array exceeds 32-item limit returns validation error
- nonpersonalMetadata not set to true causes rejection
- Invalid dataClass enum value (not 'public' or 'synthetic') returns bad request
- Malformed record objects in before/after arrays cause parse failure
- Empty or mismatched record schemas between before and after may yield inconclusive parity result

## How this service works

Compare parsed typed record projections, preserving all pins and scope while ignoring order; not raw-byte equality or proof of the underlying artifact for repository/dataset/schema/method/revision-pinned manifest Supplied-record analysis, not independent certification. Full contract at the free describe URL.

## Output

Returns a structured comparison of the before and after typed record projections, indicating whether semantic parity holds (ignoring field order), which pins and scope markers are preserved, and flagging any detected mismatches between the two record sets for the specified data class.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "after": {
   "type": "array",
   "maxItems": 32,
   "minItems": 0
  },
  "before": {
   "type": "array",
   "maxItems": 32,
   "minItems": 0
  },
  "dataClass": {
   "enum": [
    "public",
    "synthetic"
   ],
   "type": "string"
  },
  "nonpersonalMetadata": {
   "type": "boolean",
   "const": true
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/scientific-dataset-records-semantic-parity-export-ec326ff2/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from www.mahastrategies.com](https://www.zero.xyz/host/www.mahastrategies.com/llms.txt)