# Agentic Task Runner (Plan → Generate → Verify)

> Agentic Task Runner (Plan → Generate → Verify) is a paid API for AI agents from pearl-valium-stickers-upc.trycloudflare.com, paid per call via x402, $0.1/call, status unknown (last checked 2026-10-31).

Executes a full multi-step agentic pipeline — planning, generation, and verification — on a local uncensored LLM for longer-running complex tasks.

## Facts

- Endpoint: POST https://pearl-valium-stickers-upc.trycloudflare.com/v1/agent-task?utm_source=zero.xyz
- Price: $0.1/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-31
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/agentic-task-runner-plan-generate-verify-bb34bf07
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_jkefzP4oSeoY1tljD6ewq

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 agentic-task-runner-plan-generate-verify-bb34bf07 -d '<json body>'
```

Example prompt: Take this complex task and run it through the full plan-generate-verify pipeline on the uncensored local model: 'Given a dataset of sales figures, write a Python script to detect anomalies, generate a summary report, and verify the output logic is correct.'

## When to prefer this

Choose this endpoint when a task genuinely requires multiple sequential reasoning steps — planning before generating, then verifying the result — rather than a single-shot completion. It is the right choice for complex code generation, multi-part analysis, or any task where self-verification matters. Prefer it over the plain chat completion endpoint when task correctness is important and you can tolerate higher latency and higher cost ($0.1 USDC per call).

## Known failure modes

- Task too short or underspecified causing a shallow plan and poor output
- Model context window exceeded for very large inputs
- Verification step may disagree with generated output leaving an unresolved conflict
- Long-running tasks may time out on the Cloudflare tunnel
- Uncensored model may produce unexpected outputs without safety guardrails
- Network instability on the trycloudflare.com tunnel causing dropped connections

## How this service works

Complete a full agentic task (plan → generate → verify). Longer running.

## Output

Returns the complete output of a multi-step agentic run: a plan, the generated content or code, and a verification/check pass. The response includes a choices array with the final result and a usage object with token consumption details.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "usage": {
   "type": "object"
  },
  "choices": {
   "type": "array"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agentic-task-runner-plan-generate-verify-bb34bf07/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from pearl-valium-stickers-upc.trycloudflare.com](https://www.zero.xyz/host/pearl-valium-stickers-upc.trycloudflare.com/llms.txt)