StealthGPT icon

StealthGPT

Generate natural, human, publish-ready long-form content or humanize existing text with StealthGPT

Actions4

Content Generation → Start Run

AI-generated

Summary

Starts a StealthGPT content generation pipeline run (research, draft, humanize and more) from a prompt. It can optionally poll until the run finishes and return the final result.

Inputs

  • Preset — Pipeline to run: Academic, SEO (default) or Social.
  • Prompt (required) — Topic or brief to generate content from. Whitespace is trimmed, and an empty prompt raises an error.
  • Platform — Shown only for the Social preset. LinkedIn (default) or Medium.
  • Enable Fact Check — Adds a fact-checking step. Default false.
  • Enable Image Generation — Generates images for the content. Default false.
  • Wait for Completion — Default true. Polls until the run reaches a terminal status and returns the result. If false, returns the creation response immediately (with the run ID) for later use with Get Run Status.
  • Poll Interval (Seconds) — Shown when waiting. How often to check status. Default 5, minimum 1.
  • Max Wait (Seconds) — Shown when waiting. Maximum wait before timing out. Default 600, minimum 1.
  • Options — Optional collection: Idempotency Key (sent as the idempotency-key header so retries don't create duplicate runs or charges), Webhook URL (StealthGPT calls it with the final result or failure), and Webhook Secret (signs the webhook payload).

Output shape

One JSON item per input item. With waiting on, it is the final run object returned by the StealthGPT API once the run is completed. With waiting off, it is the run creation response, which includes runId.

The exact fields come from the StealthGPT API and aren't defined in the node. With waiting on, a run that ends as failed or cancelled throws an error that carries the API error message and code. A timeout also throws, and the message points to Get Run Status with the run ID. With Continue On Fail enabled, errors are returned as an item with an error field.

Examples

Example 1: Generate an SEO article and wait for the finished result in the same node

Preset=SEO, Prompt='Write a long-form article about ...', Wait for Completion=true, Max Wait=600

Example 2: Create a LinkedIn post with fact checking, fetched later

Preset=Social, Platform=LinkedIn, Enable Fact Check=true, Wait for Completion=false; pass runId to Get Run Status afterward

Example 3: Retry-safe run that notifies a webhook when done

Wait for Completion=false, Options: Idempotency Key='post-123', Webhook URL, Webhook Secret (16+ chars)

Links

Discussion