Skills/Glama/chatgpt-desktop-mcp

chatgpt-desktop-mcp@0.4.0

CAUTION· startup only

Publisher claims: "Enables other agents to list, read, start, and steer Codex chats in the macOS ChatGPT desktop app, including sending follow-ups, choosing model/reasoning/speed, interrupting work, and answering approvals. It can run locally or behind bearer-token-protected tunnels for remote monitoring and control." Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry GlamaPublisher kevinrss01 · publishing for 3 daysInstalls 0Last detonated 3d ago
What would fix thisWatchDispute verdict

Coverage

Startup only; authenticated tool behavior not tested.
Tested
server startup over stdio (initialize sent) · every file read, file written and host contacted by what ran
Not tested
the tool list (the server exited before answering) · what the tools do when called with real credentials
Needs
--tunnel-credentials — give it to the server where it runs, through vet-skills run --secrets, so the value never enters the server's environment. Never paste a credential into a chat, a ticket or a report.

What this means for the label: Caution here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.

Before you give it your token

It will ask for
CDM_TOKEN or --tunnel-credentialsConfiguration, not credentials: CDM_URL, CHATGPT_DESKTOP_MCP_CODEX_BIN, CHATGPT_DESKTOP_MCP_CONFIG_DIR, CHATGPT_DESKTOP_MCP_LOG, CODEX_HOME, TUNNEL_ORIGIN_CERT
Where it talks to
no vendor host named in the codeAlso apple.com, cloudflare.com — not the publisher's domain; check what for before connecting.
What the honeypot saw
no credential files read · no network
What we could not test
the tool list (the server exited before answering) · what the tools do when called with real credentials

Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names apple.com, cloudflare.com as well as its own vendor. The guard injects CDM_TOKEN only into requests to the host you bind (and its subdomains) and refuses the rest.

$vet-skills run --net declared -- <the server's command>

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; startup only; authenticated tool behavior not tested) on 2026-11-12 and scored it 78/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Auto-update or dynamic fetch configuration; Auto-update or dynamic fetch configuration; HTTP URL with a raw public IP address. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 24 s · 2 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
4.1sprocessdependency install exit 0
24sclaimedstartup only: exited before answering, asking for --tunnel-credentials
24sprocessscript backend runs setup only
24sprocessno install scripts to re-run

Static findings

HIGH
Auto-update or dynamic fetch configurationdist/service.js:491 · Auto-update or dynamic fetch configuration — agent may load new code at runtimeprogram: [cloudflared, 'tunnel', '--no-autoupdate', '--loglevel', 'warn', '--config', tunnelConfig, 'run', tunnelId],Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
HIGH
Auto-update or dynamic fetch configurationsrc/service.ts:549 · Auto-update or dynamic fetch configuration — agent may load new code at runtimeprogram: [cloudflared, 'tunnel', '--no-autoupdate', '--loglevel', 'warn', '--config', tunnelConfig, 'run', tunnelId],Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
HIGH
HTTP URL with a raw public IP addresssrc/service.ts:655 · HTTP URL with a raw public IP address — a hard-coded destination that domain-based monitoring cannot nameif (!publicOk) internet = await probeHealth('https://1.1.1.1/cdn-cgi/trace', 5_000);Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
URL using a raw public IP addresssrc/service.ts:655 · URL using a raw public IP address — may bypass domain-based filteringif (!publicOk) internet = await probeHealth('https://1.1.1.1/cdn-cgi/trace', 5_000);Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
URL using a raw public IP addressdist/service.js:586 · URL using a raw public IP address — may bypass domain-based filteringinternet = await probeHealth('https://1.1.1.1/cdn-cgi/trace', 5_000);Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
URL with embedded credentialstest/machines.test.ts:148 · URL with embedded credentials — may be phishing or credential leakingexpect(() => normalizeMachineUrl('https://user:[email protected]')).toThrow(/token/);Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST07
MEDIUM
File permission modification commandpackage.json:18 · File permission modification command — agent changes file ownership/permissions"build": "tsc -p tsconfig.json && chmod +x dist/cli.js",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
HTTP URL with a raw public IP addressdist/service.js:586 · HTTP URL with a raw public IP address — a hard-coded destination that domain-based monitoring cannot nameinternet = await probeHealth('https://1.1.1.1/cdn-cgi/trace', 5_000);Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07

Cost to the agent

Instructions loaded
≈42 tok
1 file · 54 in bundle
Tool list per prompt
≈0 tok
0 tools · 0 per tool
Handshake
no reply
launch to tools/list, stdio
Undocumented tools
0
description under 20 chars · avg 0

Tokens estimated at four characters each. The tool list is re-read by the agent on every turn, so it is the number to compare between servers that do the same job: ranked across the index.

Known vulnerabilities

Vulnerability advisories (OSV) cover npm and PyPI packages; this registry has no advisory database.

What would make this Verified

2 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%
  1. Unclaimed publisher≈ +8 on the totalClaim kevinrss01: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  2. Auto-update or dynamic fetch configuration≈ +4 on the totalPin versions and sources; vendor anything fetched at install time.
  3. HTTP URL with a raw public IP address≈ +4 on the totalReplace encoded or assembled strings with plain literals and remove undeclared network calls.
  4. URL using a raw public IP address≈ +2 on the totalReplace encoded or assembled strings with plain literals and remove undeclared network calls.
  5. URL with embedded credentials≈ +2 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
  6. File permission modification command≈ +2 on the totalRequest the narrowest permission that works; replace wildcards with the exact paths or hosts.
Publish a new version to Glama, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim kevinrss01Propose a patch to the maintainer