Gui AgentSAFE
GUI agent using WebMCP polyfill. Control web interfaces with natural language
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
gui-agent
An open-source, WebMCP-powered GUI agent that lives inside your web app. Let a natural-language agent operate your UI — click, fill, navigate, call your app's own actions — driven by any LLM you bring.
The optional visualizer in action: status chips per tool call, an animated gradient glow that tours every element the agent touches, and a backdrop veil that keeps the user's focus on the action.
Inspired by page-agent, but built on the emerging **WebMCP** standard via a polyfill.
- 📦 Just an npm package.
npm i @aralroca/gui-agentandimport. No script inject
d96a0a05b378OBSERVED · 2026-10-10Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add gui-agent -- npx -y @aralroca/[email protected]
{
"mcpServers": {
"gui-agent": {
"command": "npx",
"args": [
"-y",
"@aralroca/[email protected]"
]
}
}
}Exposed tools (26)
22 read · 2 write · 2 destructive. Blast radius: 2 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
abort_probe | read | probe |
abortable | read | d |
add_workflow_step | write | Add a step to the workflow on the Workflows tab (React Flow canvas). |
boom | read | d |
click | read | Click an element (button, link, checkbox, tab...) by its ref. |
dead | read | d |
delete_user | destructive | Delete |
drag | destructive | Drag an element (a card, a row, a file) and drop it onto another, as a mouse does: e.g. move a card to another column of a board. |
dup | read | d |
fill | read | Type a value into a text input, textarea, or contenteditable element by its ref. |
go_to_tab | read | Switch the console to a tab. |
invite_member | read | Invite a teammate by email with a role (Viewer, Editor, Admin). |
list_users | read | List |
n | read | d |
navigate | read | Navigate the browser to a URL. Note: this reloads the page. |
ok | read | d |
ping | read | Ping |
read_page | read | Read a text outline of the current page: interactive elements with stable refs (e.g. e7), their roles, labels, and state. Call this first, and again after the page changes. |
search | read | Search things |
search_users | read | Filter the users table by a name query. |
select_option | read | Choose an option in a <select> dropdown by its ref, matching option label or value. |
sign_in | read | Sign a user in |
t | read | d |
upload_file | write | Upload a chat attachment into a file input or dropzone on the page. Pass the attachment |
wait_for_text | read | Wait (up to ~5s) until the given text appears anywhere on the page, then re-read it. |
x | read |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (3)
delete_user, drag
@mcp-b/webmcp-polyfill, @types/react, @types/react-dom, @xyflow/react, ai, jsdom, react, react-dom
docs/demo.gif
Gates applied: no_behavioural_pass.
d96a0a05b378full audit observations/trust-audit/mcp-server/aralroca__gui-agent.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-10 | d96a0a05b378 | SAFE | B | 89 | first audit |
Questions
What is the @aralroca/gui-agent MCP server?
GUI agent using WebMCP polyfill. Control web interfaces with natural language
What tools does @aralroca/gui-agent expose?
26 in total: 22 read-only, 2 that write, and 2 that can delete or overwrite (delete_user, drag). Every one is listed on this page with its risk.
Is @aralroca/gui-agent safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 2 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does @aralroca/gui-agent need?
No credential environment variables were found in its source, so it appears to need none.
How current is this page?
The grade is for one exact copy of the source (d96a0a05b378), read on 2026-10-10. The repository is watched and re-audited when it changes.