StealthGPT

Agent Instructions

Start with https://docs.stealthgpt.ai/api-reference/quickstart.md for REST setup and https://docs.stealthgpt.ai/api-reference/models.md for Super, Standard and Lite model selection. For existing drafts, use the Writer & Humanizer endpoints; for research-to-draft generation, use the Stealth Agent guides. Choose polling, webhook callbacks or MCP with https://docs.stealthgpt.ai/api-reference/workflow-patterns.md. For bulk humanization and editorial review, use https://docs.stealthgpt.ai/api-reference/recipes/humanize-review.md. For fixed links, citations or other approved spans, use https://docs.stealthgpt.ai/api-reference/text-preservation.md and replace each protected span with a marker in your application. For assistant setup and verified client status, use https://docs.stealthgpt.ai/api-reference/integrations/mcp.md. Consult the Pricing and Data Handling references for API entitlements, word accounting and processing details; web subscriptions and API plans are separate.

Humanize text and run Stealth Agent through MCP

Connect StealthGPT to Codex, Claude and compatible clients for draft humanization and content workflows, with setup, model and billing guidance

Connect StealthGPT's remote MCP (Model Context Protocol) server to bring humanization and content workflows into your agent's conversation. Send an existing draft to humanize_text, start Stealth Agent from a brief with generate_content, and collect the result through get_run_status.

[!NOTE]

Choose a client that supports remote Streamable HTTP MCP with OAuth or bearer authentication. Client setup, tool access and account controls differ. The table below separates documented setup from a dated workflow check.

Client setup and verification

Use the setup route for your client. A documented configuration is distinct from a completed runtime check on a specific version.

Client or integration	Setup route	Verification record
Codex	OAuth setup and example prompt	November 7, 2026: codex-cli 0.167.0-alpha.3.2; OAuth sign-in and humanize_text → get_run_status completed.
Claude.ai and Claude Desktop	Custom connector with OAuth; Desktop also has a bearer-token bridge example below	Documented connector and bearer-bridge setup.
Claude Code	Client setup	Documented client setup.
Cursor	Bearer-token configuration	Documented bearer-token configuration.
Manus	Client setup	Documented client setup.
ChatGPT	Custom MCP plugin setup, subject to account and workspace controls	Setup path documented; a StealthGPT-specific end-to-end check is pending.
n8n	Community node and HTTP Request guide	Node fields checked against published n8n-nodes-stealthgpt 0.1.2 on November 7, 2026; this is a package-source check.
WordPress and Zapier	WordPress guide and Zapier guide	Dedicated integration setup guides.

Setup and verification summary reviewed November 7, 2026. Use the linked instructions for authentication and available controls.

Server endpoint

https://www.stealthgpt.ai/api/mcp/mcp

Transport: Streamable HTTP (the modern MCP transport).

Authentication

The server supports two ways to authenticate, and both resolve to the same StealthGPT account (identical billing):

• OAuth (one-click) — used by claude.ai (web) and other clients with built-in custom-connector support. You sign in with your StealthGPT account and approve access; there's no token to copy. See Connect in claude.ai (web).

• API token (bearer header) — used by IDE/desktop clients like Cursor, VS Code, Claude Desktop, and CLIs:

  Authorization: Bearer YOUR_API_TOKEN
  

  Get your token from the Stealth API dashboard.

Tools

Tool	Purpose	Key arguments
generate_content	Start a full Stealth Agent run (research → draft → optional fact-check → humanize → optional images).	preset (academic | seo | social), prompt (up to 20,000 characters), platform (for social), enableFactCheck, enableImageGeneration
humanize_text	Start a humanization run for existing text.	text (up to 200,000 characters), qualityMode (fast | quality), model (super | standard | lite, defaults to super), outputFormat (text | markdown, defaults to markdown)
get_run_status	Get the current status and final result of a run.	runId
enable_pay_as_you_go	StealthAPI accounts: get a link where you add a card to turn on pay-as-you-go billing. See Turn on pay-as-you-go from the agent.	none

generate_content and humanize_text return a runId immediately (the work runs asynchronously). The agent then calls get_run_status with that runId until the status is completed, failed, or cancelled.

The MCP server exposes the four tools listed above. For a separate AI detection request, use the REST detection endpoint.

Choose your humanization model

humanize_text uses super by default. Pass model: "standard" or model: "lite" when you want a different model, or set model: "super" explicitly to keep your workflow's choice clear.

[!NOTE]

REST humanization requests require model for accounts created after September 8, 2026 at 13:00 UTC. Older accounts retain a legacy standard default when it is omitted. MCP uses super by default regardless of account age. The deprecated heavy value maps to standard.

[!NOTE]

On a completed humanization, howLikelyToBeDetected is a 0–100 human-likeness score where higher is better (more human, less likely to be flagged as AI).

[!NOTE]

humanize_text defaults to outputFormat: "markdown", unlike the REST endpoints, which default to "text". Pass "text" explicitly if you want unformatted prose.

Install

Using Codex? Connect StealthGPT with OAuth and humanize text inside your coding session.

One-click for Cursor:

[!NOTE]

The one-click button adds the server URL; add your Authorization: Bearer YOUR_API_TOKEN header in the generated mcp.json. (Cursor connects with a bearer token. For a no-token, one-click sign-in use claude.ai (web) with OAuth — see below.)

Or configure manually for your client:

[tab] Cursor

Add to .cursor/mcp.json (project) or ~/.cursor/mcp.json (global), then reload MCP servers:

{
  "mcpServers": {
    "stealthgpt": {
      "url": "https://www.stealthgpt.ai/api/mcp/mcp",
      "headers": { "Authorization": "Bearer YOUR_API_TOKEN" }
    }
  }
}

[tab] VS Code

Add to .vscode/mcp.json:

{
  "servers": {
    "stealthgpt": {
      "type": "http",
      "url": "https://www.stealthgpt.ai/api/mcp/mcp",
      "headers": { "Authorization": "Bearer YOUR_API_TOKEN" }
    }
  }
}

[tab] Windsurf

Add to your Windsurf MCP config (~/.codeium/windsurf/mcp_config.json):

{
  "mcpServers": {
    "stealthgpt": {
      "serverUrl": "https://www.stealthgpt.ai/api/mcp/mcp",
      "headers": { "Authorization": "Bearer YOUR_API_TOKEN" }
    }
  }
}

[tab] Claude Desktop

Add it as a custom connector (Settings → Connectors → Add custom connector) using the URL https://www.stealthgpt.ai/api/mcp/mcp and sign in with OAuth — same flow as claude.ai (web).

Alternatively, bridge over stdio with a bearer token using mcp-remote:

{
  "mcpServers": {
    "stealthgpt": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://www.stealthgpt.ai/api/mcp/mcp",
        "--header",
        "Authorization: Bearer YOUR_API_TOKEN"
      ]
    }
  }
}

[tab] Other clients

Any client that supports a remote (Streamable HTTP) MCP server with custom headers:

• URL: https://www.stealthgpt.ai/api/mcp/mcp
• Header: Authorization: Bearer YOUR_API_TOKEN

Once connected, the tools appear in the client and the agent calls them when relevant — for example:

Humanize this paragraph: …

or

Generate an SEO article about home solar energy.

Example flow

1. The agent calls humanize_text with your text → receives { "runId": "…", "status": "queued" }.
2. The agent polls get_run_status with that runId.
3. When status is completed, it returns the humanized result, howLikelyToBeDetected, and billing fields (wordsSpent, creditsSpent, remainingCredits, billingMode, meteredChargedCredits).

Save the accepted run ID and retrieve that run rather than resubmitting the draft while it is processing. Route completed text through your factual, link and formatting review before delivery. For scripts or publishing queues, the TypeScript and Python review recipes show result collection, stable retry keys, signed callbacks and an editorial handoff.

Compare the available integration approaches in the AI humanizer API comparison. The StudySim API case study shows a related XYZAI product's Super quality-mode publishing workflow with original Search Console exports. The data-handling guide covers processing roles and account history.

Billing

How MCP runs are billed depends on your account:

• StealthGPT subscription — runs come out of your plan, the same as on the website. humanize_text uses your plan's words, and generate_content uses one Stealth Agent article from your monthly allowance. Completed runs report billingMode: "subscription". Connect with OAuth by signing in with your StealthGPT account.
• StealthAPI account — runs are billed in API credits, prepaid or pay-as-you-go, like the REST API.

If you have both an active subscription and a StealthAPI account, MCP runs use your API credits by default. You can switch them to your subscription under MCP billing in your usage settings. This setting only applies to MCP runs; REST API calls always use API credits.

If your plan has no words or articles left, the run is refused with a message that links to your plan.

Turn on pay-as-you-go from the agent

This applies to StealthAPI accounts. If a run fails with Payment required, the agent can call enable_pay_as_you_go. It returns a setupUrl for a Stripe page where you add a card. Pay-as-you-go turns on when you finish, and the agent can retry the run. The agent can't complete this step for you, and the link expires after 24 hours.

If pay-as-you-go is already on, the tool returns status: "active". If you don't have an API plan yet, it returns status: "plan_required" with a pricingUrl where you pick one; pay-as-you-go turns on with your first plan payment. If your last invoice is unpaid or you haven't accepted the pay-as-you-go terms, it returns a link to your usage settings instead. Accounts billed through a subscription get status: "subscription_billing". Seat participants and accounts with frozen billing get status: "not_allowed".

Connect in claude.ai (web)

The MCP server is available as a custom connector inside claude.ai — connect once with one-click OAuth (no API token to paste):

1. Open connector settings

In claude.ai, go to Settings → Connectors → Add custom connector.

2. Add the server

Set a name (e.g. StealthGPT) and the URL:

https://www.stealthgpt.ai/api/mcp/mcp

3. Sign in and approve

Click Connect, sign in with your StealthGPT account, and approve access on the consent screen. The connector switches to Connected.

4. Use the tools

Ask Claude to generate or humanize content, e.g. "Humanize this paragraph…" or "Generate an SEO article about home solar energy." Claude calls the tools and polls the run for you.

[!NOTE]

OAuth and the bearer token resolve to the same StealthGPT account, so runs are billed identically regardless of how you connect.

ChatGPT

ChatGPT supports custom MCP connections through plugins, subject to your account and workspace controls. Follow the official ChatGPT connection guide: open Plugins → Add custom MCP server, enter the public server URL, configure authentication, review the connection information and create the plugin. Complete StealthGPT sign-in where prompted, review the discovered tools, install the connection and select it with @ in a new conversation.

Use the same server URL shown above. Check that humanize_text, generate_content and get_run_status appear, then test one draft and retrieve its completed result. This is the documented setup path; a dated StealthGPT-specific ChatGPT end-to-end verification record is pending. The Codex check in the table above applies to the recorded Codex client and version.

Notes & limits

• Runs are billed to the signed-in account, as described in Billing.
• Generation can take from seconds (humanize-only, lite) to a few minutes (full academic/seo pipeline). Because it's async, the tool call returns quickly and the agent polls.
• See the async workflow and Stealth Agent async runs pages for the underlying REST contract.

This documentation is built and hosted on Mintlify, a developer documentation platform.