Skip to main content
Glama

Browser Workflow

browser_workflow
Destructive

Use this for an explicitly requested sequence of 1–12 guarded DOM actions and bounded waits on one inspected Chromium page. Validate CSS selectors first. The workflow cannot run arbitrary JavaScript, upload files, switch origins or implicitly submit a form; an explicit click may submit data or cause irreversible remote changes. Stop at the first failed step and never echo field values.

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": {
      "device": "<device>",
      "steps": "<steps>"
    },
    "name": "browser_workflow"
  }
}

Input Schema

NameRequiredDescriptionDefault
stepsYesWorkflow on one page: explicit CSS actions and bounded waits, max 12 seconds. Never silently submits; click must be explicit.
titleNoChoose a page whose title contains this text.
deviceYesReMCP device id returned by list_devices. Required for every device operation.
endpointNoLoopback Chrome DevTools endpoint. Defaults to REMCP_CDP_URL or http://127.0.0.1:9222.
target_idNoPage target id returned by browser_tabs.
timeout_msNoTimeout in milliseconds.
url_containsNoChoose a page whose URL contains this text.

Output Schema

NameRequiredDescriptionDefault
dataNoBrowser element action structured result for browser_action. Stable fields are declared explicitly; platform-specific fields may be added. Large results may return truncated/bytes/preview instead of the full payload.
textYesConcise human-readable result of the tool call.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changed
    • addedInput schema / properties / steps / items / properties / fields / items / properties / expected_value
      Added value: +{
      +  "description": "Optional exact prior field value (CAS); never returned.",
      +  "type": [
      +    "string",
      +    "boolean"
      +  ]
      +}
    • addedInput schema / properties / steps / items / properties / fields / items / properties / label
      Added value: +{
      +  "description": "Exact associated label / aria-label for one form input; use label instead of selector.",
      +  "type": "string"
      +}
    • changedInput schema / properties / steps / items / properties / fields / items / required
      Previous value: -[
      -  "selector",
      -  "value"
      -]New value: +[
      +  "value"
      +]
  2. Changed2 schema fields changed
    • addedInput schema / properties / steps / items / properties / fields / items / properties / expected_tag
      Added value: +{
      +  "description": "Optional HTML tag from Forms; rejects stale selector mismatches.",
      +  "enum": [
      +    "input",
      +    "select",
      +    "textarea"
      +  ],
      +  "type": "string"
      +}
    • addedInput schema / properties / steps / items / properties / fields / items / properties / expected_type
      Added value: +{
      +  "description": "Optional field type from Forms; checked immediately before batch writes.",
      +  "enum": [
      +    "text",
      +    "email",
      +    "password",
      +    "tel",
      +    "url",
      +    "search",
      +    "number",
      +    "date",
      +    "time",
      +    "datetime-local",
      +    "month",
      +    "week",
      +    "color",
      +    "range",
      +    "checkbox",
      +    "radio",
      +    "file",
      +    "hidden",
      +    "submit",
      +    "reset",
      +    "button",
      +    "select",
      +    "textarea"
      +  ],
      +  "type": "string"
      +}
  3. Added

TDQS

A4.5/5.0
Behavior5/5

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

Annotations already declare destructiveHint=true and openWorldHint=true, and the description builds on that with concrete consequences: an explicit click may submit data or cause irreversible remote changes, execution halts at the first failed step, field values are never echoed, and the whole workflow is bounded to roughly 12 seconds. That is real operational context an agent cannot get from the annotations.

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?

Three dense sentences with the positive use case front-loaded, followed by capability limits and then safety consequences. No filler, and each clause carries distinct decision-relevant information.

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 a destructive, multi-step batch tool, the definition covers when to use it, what it refuses to do, failure behavior, and the irreversible-change risk. With an output schema present, return values need no explanation, so nothing an agent needs before calling is missing.

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%, so every parameter and nested step field is already documented, which sets the baseline at 3. The description adds only procedural framing ('validate CSS selectors first', the 1–12 step bound, 'bounded waits'), not new meaning about individual parameters such as target_id versus title/url_contains page selection.

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 with scope: a bounded sequence of 1–12 guarded DOM actions and waits on one inspected Chromium page. The word 'sequence' plus the step count immediately separates it from the single-action siblings (browser_click, browser_fill_form, browser_wait) without needing to name them.

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

Usage Guidelines4/5

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

Gives a clear selection condition ('explicitly requested sequence') and strong exclusion boundaries: no arbitrary JavaScript, no file uploads, no origin switching, no implicit form submission — each of which implies routing to browser_evaluate, browser_upload, or browser_navigate_url. It stops short of naming those alternatives explicitly, so the routing is inferential rather than spelled out.

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.

Resources