# Release Change Low-Confidence Review Queue

> Release Change Low-Confidence Review Queue is a paid API for AI agents from market.datapackvibe.com, paid per call via x402, $0.01/call, status unknown (last checked 2026-10-15).

Identifies and surfaces low-confidence classifier predictions from a naive-Bayes release-change text classifier across up to 200 labeled records for human review.

## Facts

- Endpoint: POST https://market.datapackvibe.com/x402/assistant-sector-ml-release-change-review-queue?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/release-change-low-confidence-review-queue-442eb732
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_kDtlpagRcOZOaX4MF7r3-

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 release-change-low-confidence-review-queue-442eb732 -d '<json body>'
```

Example prompt: I've got 150 labeled release-change records with actual labels, predicted labels, and probability scores — can you run them through the low-confidence review queue to surface which ones the naive-Bayes model was most uncertain about so we can manually review them?

## When to prefer this

Choose this endpoint when you have already run a naive-Bayes or similar probabilistic classifier on release-change text and need to surface uncertain predictions for human review — particularly when your dataset is small (up to 200 records) and you want a lightweight local-tool evaluation pass without invoking a foundation model or hosted ML service. Best suited for MLOps pipelines that need a cheap ($0.01) automated triage step before manual labeling review.

## Known failure modes

- Probabilities array not summing to 1 or containing values outside [0,1] range
- Records exceeding 200-item limit rejected
- Text field exceeding 1000-character maxLength truncated or rejected
- Missing required fields (id, text, actual, predicted, probabilities) causes validation error
- Enum mismatch in actual/predicted fields (values outside feature/fix/documentation/maintenance) causes rejection
- Empty records array returns empty or trivial deliverable

## How this service works

Release change - Low-confidence review queue: Low-confidence review queue for Release change. Up to 200 supplied labelled text records. Local dataset tools, evaluation and small naive-Bayes baselines; no hosted model or foundation-model training. One of 25 profiles or calculations across 20 shared operations in this sector. 0.01 USDC includes one bounded low-confidence review queue deliverable.

## Output

A bounded deliverable report containing the subset of submitted labeled records that the classifier flagged as low-confidence, based on the supplied probability distributions, enabling targeted human review of uncertain predictions across the four release-change categories: feature, fix, documentation, and maintenance.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "records": {
   "type": "array",
   "items": {
    "type": "object",
    "required": [
     "id",
     "text",
     "actual",
     "predicted",
     "probabilities"
    ],
    "properties": {
     "id": {
      "type": "string",
      "default": "",
      "maxLength": 80,
      "minLength": 1
     },
     "text": {
      "type": "string",
      "default": "",
      "maxLength": 1000
     },
     "actual": {
      "enum": [
       "feature",
       "fix",
       "documentation",
       "maintenance"
      ],
      "type": "string"
     },
     "predicted": {
      "enum": [
       "feature",
       "fix",
       "documentation",
       "maintenance"
      ],
      "type": "string"
     },
     "probabilities": {
      "type": "array",
      "items": {
       "type": "number",
       "default": 0,
       "maximum": 1,
       "minimum": 0
      },
      "default": [],
      "maxItems": 4,
      "minItems": 4
     }
    },
    "additionalProperties": false
   },
   "default": [
    {
     "id": "example-0-0",
     "text": "feature sample document number 0 for this supplied dataset.",
     "actual": "feature",
     "predicted": "feature",
     "probabilities": [
      0.8,
      0.06666666666666667,
      0.06666666666666667,
      0.06666666666666667
     ]
    },
    {
     "id": "example-0-1",
     "text": "feature sample document number 1 for this supplied dataset.",
     "actual": "feature",
     "predicted": "feature",
     "probabilities": [
      0.8,
      0.06666666666666667,
      0.06666666666666667,
      0.06666666666666667
     ]
    },
    {
     "id": "example-0-2",
     "text": "feature sample document number 2 for this supplied dataset.",
     "actual": "feature",
     "predicted": "feature",
     "probabilities": [
      0.8,
      0.06666666666666667,
      0.06666666666666667,
      0.06666666666666667
     ]
    },
    {
     "id": "example-1-0",
     "text": "fix sample document number 0 for this supplied dataset.",
     "actual": "fix",
     "predicted": "fix",
     "probabilities": [
      0.06666666666666667,
      0.8,
      0.06666666666666667,
      0.06666666666666667
     ]
    },
    {
     "id": "example-1-1",
     "text": "fix sample document number 1 for this supplied dataset.",
     "actual": "fix",
     "predicted": "fix",
     "probabilities": [
      0.06666666666666667,
      0.8,
      0.06666666666666667,
      0.06666666666666667
     ]
    },
    {
     "id": "example-1-2",
   
… (truncated)
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/release-change-low-confidence-review-queue-442eb732/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from market.datapackvibe.com](https://www.zero.xyz/host/market.datapackvibe.com/llms.txt)