Skip to main content
Glama
Sign In

Irmu

by com.irmu.app
OverviewDiscussionsAdmin

Screenshot a page

screenshot

Take a full-page JPEG screenshot of a web page through a proxied browser. Returns the image, plus data with an AI answer if ai_query was asked. Spends a flat number of credits. If status is "queued" or "running", call crawl_poll with the job_id to collect it.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "url": "<url>"
    },
    "name": "screenshot"
  }
}

Input Schema

NameRequiredDescriptionDefault
jsNoRender the page in a browser so javascript runs. Turn off for static pages to pay less.
urlYesThe page to fetch. http or https, with a host.
cityNoThe city inside that country to appear to browse from, as the pool spells it: letters, digits, hyphens or underscores, e.g. berlin or new_york. Requires country. Doubles the price of the crawl.
touchNoReport the screen as touch capable. Requires viewport.
imagesNoDownload the images on the page.
mobileNoTell the page it is on a phone. Requires viewport; pair it with user_agent for sites that sniff it.
adblockNoDrop ads and trackers before the page loads. Free either way.
countryNoThe country to appear to browse from, as a two-letter ISO 3166-1 code such as es, us or de. Free. Required to use city.
premiumNoFetch through the premium proxy pool instead of the standard one. Costs more; use it for targets that refuse ordinary requests.
ai_queryNoA question to answer about the page once it has been fetched, e.g. "What is the price?". Costs extra credits.
viewportNoThe window size as WIDTHxHEIGHT, e.g. 390x844. Required for device_scale_factor, mobile, landscape and touch.
landscapeNoReport the screen as landscape. Requires viewport.
user_agentNoThe user agent to send instead of the crawler's own.
wait_untilNoWhen the page counts as loaded. Use domcontentloaded or load for pages that keep a connection open.networkidle2
cookies_bannerNoHow to answer cookie banners: accept opts in, reject opts out where the banner allows it, none leaves them on the page.accept
device_scale_factorNoThe device pixel ratio, e.g. 3 for a recent phone. Requires viewport.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • addedInput schema / properties / cookies_banner
      Added value: +{
+  "default": "accept",
+  "description": "How to answer cookie banners: accept opts in, reject opts out where the banner allows it, none leaves them on the page.",
+  "enum": [
+    "accept",
+    "reject",
+    "none"
+  ],
+  "type": "string"
+}
  2. Changed1 schema field changed
    • changedInput schema / properties / wait_until / default
      Previous value: -"networkidle0"New value: +"networkidle2"
  3. First observed

TDQS

A4/5.0
Behavior4/5

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

Annotations only declare openWorldHint=true and destructiveHint=false, so the description carries the real behavioral weight: it discloses the return payload (image plus `data` when `ai_query` is set), that a flat number of credits is spent, and the async status/poll contract. It stops short of noting image size limits or failure modes.

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

Conciseness5/5

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

Four short sentences, each earning its place: output format first, then the conditional extra, then cost, then the async follow-up. Nothing is padded and the most decision-relevant facts are front-loaded.

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

Completeness4/5

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

For a tool with no output schema, the description explains what comes back and how to retrieve deferred results. The remaining gap is minor: it does not characterize the screenshot's dimensions or behavior under the many viewport/device options, though those are fully documented in the schema.

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

Parameters3/5

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

Schema description coverage is 100% and the 16 parameters are richly annotated with defaults, patterns and cross-parameter requirements, so the schema does the heavy lifting. The description only adds the meaning of `ai_query` and a cost note; baseline 3 applies.

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?

States a specific verb and resource ("Take a full-page JPEG screenshot of a web page") plus the delivery mechanism (proxied browser), so the agent knows exactly what artifact it gets. The mention of `crawl_poll` for queued jobs also signals how this differs from the plain crawl sibling.

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

Usage Guidelines3/5

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

The description gives a clear conditional workflow: if `status` is "queued" or "running", call `crawl_poll` with the `job_id`. However, it never says when to choose screenshot over `crawl` or `crawl_recipe`, nor any exclusion conditions, so the core routing decision is left implicit.

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

All Irmu toolsInspect
Try in Browser
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources