Skip to main content
Glama

Zoom and Pan

zoom
Destructive

Create animated pans and magnification into recorded screen activity, then return to full view using keyframes or automatically planned from logged clicks and screen changes.

Instructions

Smooth animated zoom-and-pan into the action and back out, from keyframes or planned from logged clicks or screen changes.

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": {
      "input": "<input>"
    },
    "name": "zoom"
  }
}

Input Schema

NameRequiredDescriptionDefault
inputYesPath to the source video.
dryRunNoReturn the planned keyframes only (editable, pass back).
eventsNoEvents log (default: the input's .events.jsonl); must match the input's timing and size, so use the original recording.
followNoWithout keyframes: clicks (from the events log of start_recording events: true) or changes (where the screen changed). auto (default): clicks if logged, else changes.
outputNoOptional output path. Defaults under SCREENCAST_HOME/edits.
holdSecNoAuto mode: stay zoomed this long after activity (default 2).
maxZoomNoAuto mode: the furthest it zooms in (default 2.5).
keyframesNoViews reached by time t: a source rectangle (x, y, width, height), or { t, full: true } to zoom out. Omit to plan them (see follow).
overwriteNoAllow replacing an existing file at the output path (default false).
transitionSecNoHow long each zoom/pan eases (default 0.6).
highlightColorNoRing color (default #FF4545).
highlightClicksNoRing each logged click (default false).

Output Schema

NameRequiredDescriptionDefault
keyframesYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed10 schema fields changedv0.47.0
    • changedInput schema / properties / dryRun / description
      Previous value: -"Only return the planned keyframes (edit them and pass them back); write nothing."New value: +"Return the planned keyframes only (editable, pass back)."
    • changedInput schema / properties / events / description
      Previous value: -"Event log to read clicks from (default: the input's sibling .events.jsonl). Its times and pixels must match the input, so use the original recording, not a trimmed or scaled copy."New value: +"Events log (default: the input's .events.jsonl); must match the input's timing and size, so use the original recording."
    • changedInput schema / properties / follow / description
      Previous value: -"Without keyframes: clicks zooms toward each click in the recording's input event log (start_recording events: true), panning between clicks close together; changes zooms toward where the screen changed. auto (default) uses clicks when the log has any, else changes."New value: +"Without keyframes: clicks (from the events log of start_recording events: true) or changes (where the screen changed). auto (default): clicks if logged, else changes."
    • changedInput schema / properties / highlightClicks / description
      Previous value: -"Draw an expanding ring at each click from the event log (default false). Rings move with the zoom."New value: +"Ring each logged click (default false)."
    • changedInput schema / properties / highlightColor / description
      Previous value: -"Ring color: name, #RRGGBB, or name@alpha (default #FF4545)."New value: +"Ring color (default #FF4545)."
    • changedInput schema / properties / holdSec / description
      Previous value: -"Auto mode: stay zoomed this long after activity stops (default 2)."New value: +"Auto mode: stay zoomed this long after activity (default 2)."
    • changedInput schema / properties / keyframes / description
      Previous value: -"Views to animate between: each fills the frame with a source rectangle (x, y, width, height) by time t, or { t, full: true } to zoom out. Omit to plan them automatically (see follow)."New value: +"Views reached by time t: a source rectangle (x, y, width, height), or { t, full: true } to zoom out. Omit to plan them (see follow)."
    • removedInput schema / properties / keyframes / items / properties / full / description
      Removed value: -"Zoom back out to the whole frame."
    • removedInput schema / properties / keyframes / items / properties / t / description
      Removed value: -"Seconds: the view is fully reached at this time."
    • changedOutput schema / (root)
      Previous value: -nullNew value: +{
      +  "$schema": "http://json-schema.org/draft-07/schema#",
      +  "additionalProperties": {},
      +  "properties": {
      +    "keyframes": {
      +      "items": {},
      +      "type": "array"
      +    }
      +  },
      +  "required": [
      +    "keyframes"
      +  ],
      +  "type": "object"
      +}
  2. First observedv0.32.0

TDQS

A3.6/5.0
Behavior3/5

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

Annotations already declare destructiveHint=true, idempotentHint=false, and openWorldHint=false, so the agent knows this writes something non-idempotently. The description adds that zoom is directional (in and back out) and keyframe/tracking driven, which is useful context. It does not state that a new file is rendered to `output`, nor that the source video is left untouched, which are the real behavioral questions for a mutating render tool.

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?

A single front-loaded sentence with no filler; the core behavior and the two sourcing modes are packed in without redundancy. Nothing could be cut without losing information.

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?

An output schema exists, so return values needn't be explained, and the 12 parameters are fully documented in-schema. What's mildly under-served is the write/render lifecycle: that the tool always produces an output file (defaulting under SCREENCAST_HOME/edits) and that `dryRun` is the safe preview path. For a destructive render tool, one sentence on that would make the definition complete.

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% across all 12 parameters, including enums, bounds, and the nested keyframe shape, so the schema does the heavy lifting. The description adds only the conceptual pairing of 'keyframes' vs 'planned from logged clicks or screen changes', which the schema already encodes via `follow`. Baseline 3 is appropriate.

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?

The description names a concrete verb pair (zoom-and-pan) and its effect (smooth animated, into the action and back out), plus the two input modes (keyframes or planned). It's clear what the tool produces, but it never distinguishes itself from adjacent transform tools like crop, reframe, or scale, which an agent could plausibly reach for instead.

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 phrase 'from keyframes or planned from logged clicks or screen changes' implies the two operating modes, but which mode to pick and when is left to the schema's `follow` and `keyframes` fields. There is no explicit statement of when NOT to use this (e.g. prefer reframe for changing aspect ratio, crop for static framing), so usage is only implied.

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