# Batcave Evidence Judge

> Batcave Evidence Judge is a paid API for AI agents from playstation-newbie-mary-enhancement.trycloudflare.com, paid per call via x402, $0.02/call, status unknown (last checked 2026-09-15).

Independently reviews a proposed agent action against provided evidence and returns a structured pass/caution/reject verdict

## Facts

- Endpoint: POST https://playstation-newbie-mary-enhancement.trycloudflare.com/v1/review-decision
- Price: $0.02/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/batcave-evidence-judge-212e200b
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_yggwsJjX7h3CZl9rYkFbT

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 batcave-evidence-judge-212e200b -d '<json body>'
```

Example prompt: Before you actually send that refund of $500 to the customer, run an independent evidence review — the risk level is high, here's my reasoning summary and the 3 evidence items I found — and only proceed if you get a pass verdict.

## When to prefer this

Choose this endpoint when you need an independent, evidence-bound safety gate before an autonomous agent executes a consequential or irreversible action, especially at medium or high risk levels. It is particularly valuable in agentic pipelines where a second-opinion check is required to prevent hallucinated reasoning from driving real-world actions. Prefer it over internal self-review when you need an externally auditable, structured pass/caution/reject record.

## Known failure modes

- Empty or missing proposed_action field returns validation error
- Evidence array exceeding 16 items is rejected
- Reasoning summary exceeding 4000 characters is truncated or rejected
- Invalid risk enum value causes schema validation failure
- Ambiguous or contradictory evidence may produce cautious rather than definitive verdicts
- Network timeout on Cloudflare tunnel may cause intermittent failures
- Payment not received results in 402 response before review is processed

## How this service works

Batcave Evidence Judge: independent evidence-bound review of a proposed agent action with pass/caution/reject structured output.

## Output

A structured verdict of pass, caution, or reject indicating whether the proposed action is sufficiently supported by the provided evidence, along with structured output suitable for automated decision gating in an agent pipeline.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "risk": {
   "enum": [
    "low",
    "medium",
    "high"
   ],
   "type": "string"
  },
  "task": {
   "enum": [
    "review_decision"
   ],
   "type": "string"
  },
  "job_id": {
   "type": "string",
   "maxLength": 128
  },
  "evidence": {
   "type": "array",
   "items": {
    "type": "string",
    "maxLength": 1000
   },
   "maxItems": 16
  },
  "proposed_action": {
   "type": "string",
   "maxLength": 2000,
   "minLength": 1
  },
  "reasoning_summary": {
   "type": "string",
   "maxLength": 4000
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/batcave-evidence-judge-212e200b/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from playstation-newbie-mary-enhancement.trycloudflare.com](https://www.zero.xyz/host/playstation-newbie-mary-enhancement.trycloudflare.com/llms.txt)