Skip to main content
Glama

Create a demo from web pages or screenshots

create_demo

Makes a new draft demo without the Chrome extension, from either:

  • urls: 1–20 public web pages (http or https). AtriumKit opens each in its own headless browser (desktop 1440×900 or a 390 px phone), closes an obvious cookie banner, blurs emails and phone numbers, and captures a screenshot plus the page’s HTML (editable like a recording). Each page becomes a screen, in order. Pages behind a sign-in are refused ("needs a sign-in"): record those with the extension. Capturing takes about 5–10 seconds a page; this tool waits up to wait seconds, then answers with the progress.

  • images: 1–30 screenshots, each an https url of a PNG/JPEG/WebP or base64 data (the whole request must stay under 2 MB, so prefer URLs for anything but small images). Guided demos get an intro, one tooltip per screen and an outro; clickable demos link the pages through their own navigation. Then, unless ai.copy is false, AI writes the copy (and, with ai.voice, narration and voice) in the given tone and language toward goal (AI credits; voice is Pro and Team). Answers with demoId, editUrl, the capture and AI progress, and what to do next (get_demo_progress while it runs). The demo is a draft: publish_demo puts it live.

Requires: demos:read, demos:write.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "create_demo"
  }
}

Input Schema

NameRequiredDescriptionDefault
aiNoAI copy and voice in one pass after the demo is made. Leave out for copy in a friendly tone, in English, without voice.
typeNoguided: tooltips along one path; clickable: hotspots between screens.guided
urlsNoPublic web pages, one screen each, in order, e.g. ["https://acme.com", "https://acme.com/pricing"].
waitNourls only: seconds to wait for the pages before answering (0–40).
titleNoDemo title. Default: from the first page (its site name) or the first image.
deviceNourls only: the browser’s viewport, desktop (1440×900) or phone (390×844).desktop
folderNoA folder id or name (list_workspaces).
imagesNoScreenshots, one screen each, in order. Give `urls` or `images`, not both.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A4.8/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Goes well beyond the annotations (readOnly=false, destructive=false, openWorld=true, idempotent=false) by disclosing the actual side effects: headless browser capture, cookie-banner dismissal, email/phone blurring, HTML captured as editable, 5-10s per page, a hard `wait` cutoff that returns partial progress, a 2 MB request ceiling, and AI credit/Pro-Team costs. It also states the artifact is a draft requiring publish_demo.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loads the verb and the two-mode choice, then uses bullets to keep dense constraints scannable; the Requires line is placed last where it belongs. It is long, but nearly every clause carries operational information (rate, size cap, blurring, credit costs) rather than filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For an 8-param tool with nested objects and no output schema, the description supplies return shape (demoId, editUrl, capture and AI progress), required scopes, async polling guidance, and the draft-to-publish lifecycle. Nothing an agent needs in order to call it correctly appears missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description adds cross-parameter meaning the schema doesn't: per-page ordering semantics, what `wait` actually does when exceeded (answers with progress), and what ai.copy=false yields. It stops short of explaining `type` (guided vs clickable) or `folder`, which are left entirely to the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Opens with a specific verb+resource and scope: "Makes a new draft demo without the Chrome extension, from either `urls` ... or `images`". The two input modes are named immediately, and "draft" plus the publish_demo pointer distinguish it from a recording/extension capture flow.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly splits the two usage paths and their constraints: 1-20 public http/https pages, or 1-30 screenshots, and states the exclusion "Give `urls` or `images`, not both" along with the fallback for sign-in pages ("record those with the extension"). It also names the follow-up tools in context (get_demo_progress while running, publish_demo to go live).

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.