K

kimi-agent-desktop

a-g-menshikov/agent-desktop · v0.2.0 · Kimi Plugins

by a.g.menshikovOwner verifiedScan grade: Good

Computer Use on macOS for Kimi Code and Kimi Code Desktop: AX-tree with element refs, headless clicks, typing, scrolling and screenshots through 21 MCP tools with zero dependencies, as a drop-in kimi-cu replacement

Follow0
Protect this install with HOL Guard

Guard checks the skills and MCP servers a plugin adds on your machine and asks you before a risky action runs. The scores above are free, and no plan changes them.

Is kimi-agent-desktop a fit?

From the plugin's manifest and HOL's scan. Anything the plugin does not declare is shown as not declared.

Works with

  • Kimi Code
  • MCP
  • 1 skill
  • 1 MCP server

Before you install

Sign-in
Not declared
Declared access
Not declared
License
MIT

Check before relying on it

What is kimi-agent-desktop?

kimi-agent-desktop is a published kimi plugins plugin for AI coding agents that works with Kimi Code and MCP, developed by a.g.menshikov and distributed through the HOL AI plugin registry. Computer Use on macOS for Kimi Code and Kimi Code Desktop: AX-tree with element refs, headless clicks, typing, scrolling and screenshots through 21 MCP tools with zero dependencies, as a drop-in kimi-cu replacement.

Canonical slug
a-g-menshikov/agent-desktop
Version
v0.2.0 · updated Nov 12, 2026

Registry snapshot

Owner verification
Verified on HOL as @mnshkv
Scanner
HOL plugin scanner
Security result
Needs review
Digest verified
Yes
1 bundled skillRead, copy or download each SKILL.mdOpen skills

Trust and security

Trust & reputation

On-Chain Verified
90

HOL Trust Score

Per-metric points (0–100 each) combined via a weighted average into the overall score.

Installability
100/100
Maintenance
65/100
MCP server setup
100/100
Plugin Security
89/100
Provenance
80/100
Publisher Quality
100/100

Provenance

Plugin root
.
Source repo
https://github.com/mnshkv/kimi-agent-desktop
Source commit
a7e9032ccd70…
Publisher verified
Yes
Owner verified
@mnshkv
Agent Plugins 1.0
Not detected

Continuous scanner CI detected

No action is required. This plugin receives the full trust score.

README badge not detected

Plugin owners can add the HOL badge to the repository README from Guard Plugins. Distribution, not security evidence. It adds a 2% engagement adjustment to the overall trust score and never changes the security result.

Needs review
Security result
89
Security score
0
High findings
Provider
codex-plugin-scanner
Grade
B · Good
Version
3.24.2
codex:hol-plugin-scanner-HBuP9g / cisco-skill-scanner: failedcodex:hol-plugin-scanner-HBuP9g / cisco-mcp-scanner: skipped

[codex:.] Manifest Validation

21/31

9 checks

[codex:.] Security

16/16

9 checks

[codex:.] Operational Security

N/A

5 checks · not applicable

[codex:.] Best Practices

9/12

5 checks

[codex:.] Marketplace

N/A

3 checks · not applicable

[codex:.] Skill Security

5/5

4 checks

[codex:.] Code Quality

10/10

2 checks

[kimi:.] Kimi Plugin

45/45

10 checks

Published advisories

None matched in GitHub security advisories, checked . This covers published advisories only, not problems nobody has reported.

GitHub security advisories are those the source repository has published.

Findings

infomanifest-validationPLUGIN_JSON_RECOMMENDED_AUTHOR

The manifest is missing the documented recommended field "author".

Add "author" to strengthen the plugin manifest metadata.

infomanifest-validationPLUGIN_JSON_RECOMMENDED_HOMEPAGE

The manifest is missing the documented recommended field "homepage".

Add "homepage" to strengthen the plugin manifest metadata.

infomanifest-validationPLUGIN_JSON_INTERFACE_CAPABILITIES

The interface object is missing a valid "capabilities" value.

Add a valid "capabilities" field to the interface metadata.

infomanifest-validationPLUGIN_JSON_INTERFACE_ASSET_WEBSITEURL

The interface field "websiteURL" must use HTTPS or point to a safe in-repo asset.

Update "websiteURL" to use HTTPS or an existing relative asset path.

infomanifest-validationPLUGIN_JSON_INTERFACE_ASSET_PRIVACYPOLICYURL

The interface field "privacyPolicyURL" must use HTTPS or point to a safe in-repo asset.

Update "privacyPolicyURL" to use HTTPS or an existing relative asset path.

infomanifest-validationPLUGIN_JSON_INTERFACE_ASSET_TERMSOFSERVICEURL

The interface field "termsOfServiceURL" must use HTTPS or point to a safe in-repo asset.

Update "termsOfServiceURL" to use HTTPS or an existing relative asset path.

infomanifest-validationPLUGIN_JSON_INTERFACE_ASSET_COMPOSERICON

The interface field "composerIcon" must use HTTPS or point to a safe in-repo asset.

Update "composerIcon" to use HTTPS or an existing relative asset path.

infomanifest-validationPLUGIN_JSON_INTERFACE_ASSET_LOGO

The interface field "logo" must use HTTPS or point to a safe in-repo asset.

Update "logo" to use HTTPS or an existing relative asset path.

infomanifest-validationPLUGIN_JSON_INTERFACE_ASSET_SCREENSHOTS

The interface field "screenshots" must use HTTPS or point to a safe in-repo asset.

Update "screenshots" to use HTTPS or an existing relative asset path.

infobest-practicesCODEXIGNORE_MISSING

A .codexignore file helps prevent accidental inclusion of local artifacts and secrets.

Add a .codexignore file with generated assets, local state, and secret paths.

Install guidance

Install in MCP

Connect the bundled MCP server shipped in the repository .mcp.json file.

MCP docs
  1. 1

    Clone the repository

    shell
  2. 2

    Register the server in your MCP client

    Claude Code reads a project .mcp.json file and asks you to approve its servers when you open the folder. Other clients keep their own MCP configuration, so copy the server entry from .mcp.json into that client. If this repository configures its server differently, follow the MCP setup the publisher documents.

  3. 3

    Verify the server

    Check the MCP server list in your client. The server declared in .mcp.json should report as connected. In Claude Code, run /mcp to see it.

Did setup work in MCP?

No setup reports for MCP yet. Self-reported by signed-in visitors. Copying a command is not counted, and most installs are never reported.

Attested installs: Fewer than 10 attested installs of this plugin in week 2026-W46, so no number is shown. Counted from opt-in HOL Guard install reports, once per machine and runtime, and never mixed with the answers above. How install reports work.

Loading reviews

Plugin manifest

{
  "$schema": "https://kimi.com/schemas/kimi.plugin.schema.json",
  "name": "agent-desktop",
  "version": "0.2.0",
  "description": "Computer Use для macOS/Windows/Linux поверх родного MCP-сервера agent-desktop (vercel-labs): каждая команда CLI — отдельный инструмент desktop_* с генерируемой схемой. Headless по умолчанию.",
  "keywords": [
    "computer-use",
    "macos",
    "windows",
    "linux",
    "accessibility",
    "automation",
    "screenshot",
    "gui",
    "mcp"
  ],
  "author": "a.g.menshikov",
  "license": "MIT",
  "skills": "./",
  "interface": {
    "displayName": "Agent Desktop",
    "developerName": "a.g.menshikov",
    "shortDescription": "Управление GUI через accessibility-дерево — родной MCP-сервер agent-desktop (vercel-labs)",
    "longDescription": "Плагин запускает родной MCP-сервер vercel-labs/agent-desktop (`agent-desktop mcp`): каждая команда CLI становится инструментом desktop_* с автогенерируемой схемой, каталог инструментов следует за установленным CLI без обновления плагина. Наблюдение: снапшоты accessibility-дерева с квалифицированными refs, скриншоты как image content. Действия: клики, ввод, set-value, клавиши, скролл, выбор, toggle, drag, clipboard, окна и приложения, уведомления, batch и wait, плюс песочницы (macOS VM). Headless по умолчанию: курсор не двигается, фокус не крадётся; физический ввод включается флагом --headed при старте сервера.\n\nТребуется: CLI agent-desktop >= 1.2.0 (npm i -g agent-desktop или brew install agent-desktop, см. bin/setup.sh) и права Accessibility + Screen Recording для приложения Kimi Code.",
    "category": "PRODUCTIVITY",
    "hostKind": "local"
  },
  "mcpServers": "./.registry/mcp.json",
  "platforms": [
    "macos",
    "windows",
    "linux"
  ],
  "repository": "https://github.com/mnshkv/kimi-agent-desktop",
  "holManifestOrigin": "repository",
  "registryIndexVersion": 9
}

Marketplace source

Repo URL
https://github.com/mnshkv/kimi-agent-desktop
Marketplace path
Unknown
Source path
.
Install policy
Unspecified

Skills

1 SKILL.md file ship with this plugin. Preview, copy, or download each one, then install them with the Skills CLI.

Share
skills-cli
  • agent-desktop

    skills/agent-desktop/SKILL.md

    |

    SKILL.md content is not in this snapshot. Open the raw file to copy it from the source repository.

File inventory

.codex-plugin/plugin.json

plugin-manifest

2,239 bytes

f6a14bc3b7b5eb6b…

.registry/mcp.json

mcp-config

80 bytes

dc7cce18f839faa4…

kimi.plugin.json

file

2,376 bytes

0d0c9e3c5239aef5…

LICENSE

file

1,428 bytes

e5e532a8cd663d41…

README.md

file

4,131 bytes

61d9f1896b9da3f7…

SECURITY.md

file

1,070 bytes

03adbb5c3a42ba13…

skills/agent-desktop/SKILL.md

skill

6,007 bytes

5bc71440d580a4e4…

kimi-agent-desktop — Frequently asked questions

What is kimi-agent-desktop?
kimi-agent-desktop is an AI plugin in the HOL registry. Computer Use on macOS for Kimi Code and Kimi Code Desktop: AX-tree with element refs, headless clicks, typing, scrolling and screenshots through 21 MCP tools with zero dependencies, as a drop-in kimi-cu replacement.
How do I install kimi-agent-desktop?
Install kimi-agent-desktop in your harness: MCP — git clone https://github.com/mnshkv/kimi-agent-desktop; Any agent — npx [email protected] add mnshkv/kimi-agent-desktop; Kimi Code — /plugins install https://github.com/mnshkv/kimi-agent-desktop. Full step-by-step guidance is on the HOL plugin page.
How do I install kimi-agent-desktop in MCP?
To install kimi-agent-desktop in MCP, start with git clone https://github.com/mnshkv/kimi-agent-desktop. The complete step-by-step install guide for MCP is on the HOL plugin page.
How do I install kimi-agent-desktop in Kimi Code?
To install kimi-agent-desktop in Kimi Code, start with /plugins install https://github.com/mnshkv/kimi-agent-desktop. The complete step-by-step install guide for Kimi Code is on the HOL plugin page.
How is the Trust Score for kimi-agent-desktop computed?
kimi-agent-desktop has a HOL Trust Score of 90 out of 100. The score is a weighted sum of six components from a static scan of the source, capped for high-severity findings and adjusted by two disclosed policies. Memberships and payments are not inputs, and a scan is not a safety guarantee. Full breakdown: https://hol.org/registry/plugins/a-g-menshikov%2Fagent-desktop/trust
Who publishes kimi-agent-desktop?
kimi-agent-desktop is published by a.g.menshikov and listed on HOL.

Directory acceptance

Listed in 0 of 5 directory lists

Read from the list each directory publishes on GitHub. Being listed means the directory accepted the plugin; it is not a HOL security rating.

Lists last read . HOL re-reads them every day.

Submitting it? Check it against each directory's rules

Install signal

HOL cannot see every install. These are separate signals from different sources, so they are not added together. Counts under each floor show as <N.

Install commands issued here
<5
HOL Guard install commands issued from this listing, last 28 days.
Opt-in install reports
<10
Machines that reported installing it through HOL Guard in week 2026-W46. Reporting is opt-in, so most installs are not counted.
HOL Guard workspaces
<5
Workspaces that approved it or include it in an approved-plugin feed, as of today.
New GitHub stars
<5
From HOL's GitHub snapshots of the repository, 2026-11-11 to 2026-11-22.

Reviewed comparisons

Reviews and similar plugins