Skip to main content
Glama

act_fast

Destructive

Perform a single Playwright browser action—click, fill, select, scroll, or wait—on a ref, selector, text, or x/y coordinate; optionally return a DOM snapshot and block submits unless allowSubmit=true.

Instructions

One action on a ref, selector, text or x/y. Supports returnSnapshot; blocks submit/publish/send unless allowSubmit=true.

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": {
      "action": "<action>"
    },
    "name": "act_fast"
  }
}

Input Schema

NameRequiredDescriptionDefault
xNoViewport x coordinate (or offset relative to target).
yNoViewport y coordinate (or offset relative to target).
dxNo
dyNo
msNo
keyNo
refNo
urlNoWait until page URL matches pattern.
fileNo
textNoVisible text. For wait: "A|B" matches any alternative (result.matched).
filesNo
labelNo
stateNo
tabIdNo
valueNo
actionYes
buttonNoleft
slowlyNoWhen false fills immediately; when true types character by character.
submitNoPress Enter after type action (blocked unless allowSubmit=true).
valuesNoMultiple option values/labels for select action.
delayMsNo
textNewNoWait for text ("A|B") appearing after the step starts; existing matches ignored.
dropDataNoCustom MIME type -> string data entries for drop action DataTransfer.
frameIdxNo
selectorNo
textGoneNoWait until matching text disappears.
loadStateNo
modifiersNo
targetRefNoDrop target ref for drag action.
timeoutMsNo
targetTextNoDrop target text for drag action.
allowSubmitNo
snapshotModeNo
snapshotQueryNo
returnSnapshotNo
targetFrameIdxNo
targetSelectorNoDrop target selector for drag action.
observeAftermathNoObserve click navigation and mutating requests (up to 800 ms); report possible submits.
snapshotMaxCharsNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed4 schema fields changedv0.6.1
    • addedInput schema / properties / observeAftermath
      Added value: +{
      +  "description": "Observe click navigation and mutating requests (up to 800 ms); report possible submits.",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / snapshotQuery
      Added value: +{
      +  "type": "string"
      +}
    • addedInput schema / properties / text / description
      Added value: +"Visible text. For wait: \"A|B\" matches any alternative (result.matched)."
    • addedInput schema / properties / textNew
      Added value: +{
      +  "description": "Wait for text (\"A|B\") appearing after the step starts; existing matches ignored.",
      +  "type": "string"
      +}
  2. First observedv0.4.0

TDQS

C2.9/5.0
Behavior3/5

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

Annotations already declare destructiveHint=true, readOnlyHint=false, idempotentHint=false and openWorldHint=true, so the risk profile is covered. The description adds two genuinely non-derivable facts: that returnSnapshot can attach a page snapshot, and that certain submit-like actions are hard-blocked unless allowSubmit=true. It still says nothing about timeouts, observation windows (observeAftermath), or what happens on failure for a 39-parameter destructive tool.

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?

Two short sentences, front-loaded with the action model and followed by the two most consequential flags; no filler. It is dense to the point of being cryptic ("text or x/y"), but it is not padded or repetitive.

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

Completeness2/5

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

This is the most complex tool in the sibling set — 39 parameters, 16 action variants, nested dropData, no output schema, and destructive semantics. A two-sentence description that never enumerates the action vocabulary or explains the snapshot/observeAftermath returns leaves large gaps an agent must fill by reading the raw schema.

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

Parameters2/5

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

Schema description coverage is only 36%, so the description carries real weight here, yet it only gestures at targets (ref/selector/text/x,y), returnSnapshot and allowSubmit. Roughly two dozen parameters — dx/dy, ms, key, files, label, state, values, delayMs, loadState, modifiers, targetRef/targetText/targetSelector, timeoutMs, snapshotMode/Query/MaxChars, frameIdx, dropData — are never contextualized, and several of those have no schema description either.

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

Purpose4/5

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

"One action on a ref, selector, text or x/y" gives a specific verb (perform a single page action) and enumerates the four target-resolution modes, which is meaningfully more than a restatement of the name. It does not, however, contrast itself with siblings like batch_actions or smart_fill_form, so the agent must infer that this is the single-step primitive.

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

Usage Guidelines2/5

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

The only conditional guidance is the submit/publish/send gate, which is a safety constraint rather than routing advice. Nothing says when to prefer this over batch_actions (many actions), smart_fill_form (form filling) or wait_for_human (blocking waits), even though those are exactly the decisions an agent faces here.

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