Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →dmmulroy/cua-driver
Drive a native GUI app (macOS, Windows, Linux) via the cua-driver CLI (default) or MCP server; snapshot its accessibility tree, act through snapshot-bound element tokens, native menu paths, exact window geometry, or pixel coordinates, and verify from fresh state.
Passed 16 of 19 test cases with the skill, and 6 of 19 without it.
and 14 more in the full report
11 fixed5 passed either way2 still failing1 broke3 errored
Tokens, turns and time describe resource use, not a dollar price.
One of 11 test cases the skill fixed. Same model, same prompt, with and without the skill.
An agent is in standard window scope and wants to execute press_key with key 'enter'. The agent omits the pid parameter from the tool call. Will the tool execute successfully in window scope?
| Checked | Without skill | With skill |
|---|---|---|
| States that pid is required for window-scope keyboard actions. | Missed | Passed |
| Explains that omitting pid targets the active foreground application in desktop scope. | Missed | Passed |
The instructions your model reads when the skill loads.
Read LOCAL.md before the upstream guidance below. This setup uses
the persistent computer MCP server for GUI workflows. LOCAL.md defines our
Helium profile rules and overrides upstream transport defaults. Installed tool
schemas take precedence over stale examples.
Orchestrates cross-platform app automation via cua-driver. Whenever
a user asks to drive a native app, follow the loop in this skill
rather than calling tools ad-hoc — the snapshot-before-action
invariant is not optional and silently breaks if you skip it.
When both history_status and history_query are advertised and the user asks
to continue, resume, or recall prior Cua work, call history_status first. If
history is healthy and access is admitted, make one bounded initial
history_query before broad application or window discovery. Treat returned
metadata only as a lead and verify current state through the least intrusive
appropriate source. Content, geometry, arguments, results, and user intent
omitted from the metadata remain unknown.
Make another bounded query only when the initial slice exposes a relevant session or sequence boundary; never broaden a query to reconstruct excluded fields.
Continue without history when either tool is absent, access is denied, the query is empty, or history is unhealthy. Do not query history for unrelated tasks merely because the tools are advertised, and never mutate history lifecycle or settings.
This file is the cross-platform core: snapshot invariant, CLI vs MCP choice, tool surface naming, behavior matrix, canonical loop, pixel-click contract, common failure modes. The platform-specific material (forbidden-list, accessibility tree implementation, launch semantics, click dispatch) lives in companion files in this same
test cases passed, vs 6 / 19 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.