# Should I Ask The Human?

> Should I Ask The Human? is a paid API for AI agents from agent-product-normalizer.vercel.app, paid per call via x402, $0.003/call, status unknown (last checked 2026-09-18).

Decides whether an AI agent should pause and ask the human for clarification or safely infer and proceed autonomously

## Facts

- Endpoint: GET https://agent-product-normalizer.vercel.app/api/v1/should-ask-human
- Price: $0.003/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-18
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/should-i-ask-the-human-73dfa0db
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_pfky7q2f0thZB2sP_dx4l

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 should-i-ask-the-human-73dfa0db
```

Example prompt: I'm about to send the order confirmation email to the customer, but I'm not 100% sure which shipping address to use — the one from their profile or the one they mentioned in chat. Should I ask them or just pick the profile address?

## When to prefer this

Use this endpoint when building agentic pipelines that need to decide autonomously whether to interrupt the user or self-resolve ambiguity. It is particularly valuable before irreversible actions (sending messages, placing orders, deleting data) where a wrong assumption has real consequences. Prefer this over hardcoded thresholds or ad-hoc logic when the cost of unnecessary interruptions (user friction) and the cost of wrong inference (mistakes) both matter and need to be balanced dynamically.

## Known failure modes

- Missing required 'task' query parameter returns a validation error
- Extremely long task descriptions exceeding 10,000 characters are rejected
- Ambiguous or near-empty context may result in a conservative 'ask human' recommendation even when inference might be acceptable
- Network or service availability issues on the Vercel deployment may return 5xx errors

## How this service works

Decide whether an agent should ask the human or safely infer and continue

## Output

Returns a clear recommendation on whether the agent should ask the human for clarification or safely proceed with the proposed assumption, along with supporting reasoning that explains the decision based on the task context and any assumptions the agent wanted to make.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "$schema": "https://json-schema.org/draft/2020-12/schema",
 "required": [
  "input"
 ],
 "properties": {
  "input": {
   "type": "object",
   "required": [
    "type",
    "method"
   ],
   "properties": {
    "type": {
     "type": "string",
     "const": "http"
    },
    "method": {
     "enum": [
      "GET"
     ],
     "type": "string"
    },
    "queryParams": {
     "type": "object",
     "required": [
      "task"
     ],
     "properties": {
      "task": {
       "type": "string",
       "maxLength": 10000,
       "description": "Task the agent is about to execute"
      },
      "known_context": {
       "type": "string",
       "maxLength": 15000,
       "description": "Relevant known context"
      },
      "proposed_assumption": {
       "type": "string",
       "maxLength": 4000,
       "description": "Assumption the agent wants to make"
      }
     },
     "additionalProperties": false
    }
   },
   "additionalProperties": false
  },
  "output": {
   "type": "object",
   "required": [
    "type"
   ],
   "properties": {
    "type": {
     "type": "string"
    },
    "example": {
     "type": "object"
    }
   }
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/should-i-ask-the-human-73dfa0db/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agent-product-normalizer.vercel.app](https://www.zero.xyz/host/agent-product-normalizer.vercel.app/llms.txt)