# Deterministic A2A Broken-JSON Repair & Schema Validator

> Deterministic A2A Broken-JSON Repair & Schema Validator is a paid API for AI agents from preflight.agentdiet.dev, paid per call via MPP, $0.002/call, status unknown (last checked 2026-10-21).

Deterministically repairs malformed, truncated, or markdown-wrapped JSON strings from LLM outputs so multi-agent pipelines can safely parse and continue processing.

## Facts

- Endpoint: GET https://preflight.agentdiet.dev/v1/schema-repair?utm_source=zero.xyz
- Price: $0.002/call
- Payment: MPP
- Status: unknown
- Last checked: 2026-10-21
- Activations on Zero: 0
- Provider: preflight.agentdiet.dev
- Website: https://preflight.agentdiet.dev
- Canonical page: https://www.zero.xyz/c/preflight-agentdiet-dev-deterministic-a2a-broken-json-repair-schema-c2871b4a
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_0BLcN-o6K71AbbLWUDpFE

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 preflight-agentdiet-dev-deterministic-a2a-broken-json-repair-schema-c2871b4a
```

Example prompt: This JSON from my LLM came back malformed — it has trailing commas, single quotes, and a markdown code fence around it. Can you repair it so it parses correctly: ```{'name': 'Alice', 'age': 30,}```?

## When to prefer this

Choose this endpoint when an LLM or agent in your pipeline is producing JSON with common syntax errors (trailing commas, markdown fences, single quotes, truncation) and you need a deterministic, non-LLM fix that won't introduce hallucinated fields. Prefer it over asking another LLM to re-generate or fix the JSON, since this is cheaper, faster, and reproducible.

## Known failure modes

- Input is so severely truncated or corrupted that no deterministic repair is possible
- Input is not JSON at all (e.g. plain prose), resulting in an error or empty object
- Ambiguous structure where multiple valid repairs exist and the heuristic picks the wrong one
- Payment authorization fails (x402), blocking the request
- Query string encoding issues when passing complex JSON as a GET parameter

## How this service works

Repairs truncated or malformed LLM JSON outputs (trailing commas, markdown fences, single quotes) so multi-agent pipelines never crash. (Live x402 v2, IETF MPP & MCP endpoint at https://preflight.agentdiet.dev/v1/schema-repair)

## Output

Returns a repaired, valid JSON object derived from the malformed input string, with syntax errors like trailing commas, single quotes, markdown fences, and truncated structures corrected so the output is immediately parseable.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "raw_json"
 ],
 "properties": {
  "raw_json": {
   "type": "string",
   "minLength": 1,
   "description": "Malformed, truncated, or markdown-wrapped JSON string to deterministically repair"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object"
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/preflight-agentdiet-dev-deterministic-a2a-broken-json-repair-schema-c2871b4a/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from preflight.agentdiet.dev](https://www.zero.xyz/host/preflight.agentdiet.dev/llms.txt)