Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Cloze
v1.0.2Cloze integration. Manage Organizations. Use when the user wants to interact with Cloze data.
⭐ 0· 227·0 current·0 all-time
byMembrane Dev@membranedev
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The name/description (Cloze integration via Membrane) matches the SKILL.md intent, but the skill metadata lists no required binaries or environment; the SKILL.md explicitly requires the Membrane CLI (npm package), network access, and a Membrane account. The metadata should have declared at least 'node/npm' or a required binary and network/credential requirements.
Instruction Scope
The SKILL.md stays within the stated integration purpose: it tells the agent to install and use the Membrane CLI, create/connect connectors, run actions, and optionally proxy raw requests to the Cloze API. It does not instruct reading unrelated local files or harvesting unrelated environment variables. However, it does assume/require interactive auth flows and local credential storage (membrane login), which can create local tokens the user should understand.
Install Mechanism
There is no declared install spec in the registry entry, but the README tells users to run 'npm install -g @membranehq/cli'. Installing a global npm package is a moderate-risk install mechanism (public registry) and should have been declared. The absence of an explicit install declaration and lack of verification (no pinned release, no checksum) is a discrepancy to note.
Credentials
requires.env and primary credential are empty in metadata, yet the SKILL.md requires a Membrane account and browser-based login; it also implicitly requires node/npm and network access. The skill requests access to a user's Membrane account via CLI-authenticated sessions (local tokens), but does not declare or justify any credential/environment variables in the metadata.
Persistence & Privilege
always is false and the skill is user-invocable with normal autonomous invocation allowed. The skill does not request elevated platform-wide privileges nor direct modification of other skills' configs. Note: agent-autonomous invocation combined with network-capable CLI could perform actions on the user's Cloze/Membrane account if invoked without user confirmation—this is expected behavior for integrations but worth awareness.
What to consider before installing
This skill is instruction-only and tells you to install the public npm package '@membranehq/cli' and to sign in to your Membrane account (browser auth). Before installing or invoking it: 1) Verify you trust the '@membranehq/cli' package and its publisher (check the npm page and GitHub repo) and consider installing it in an isolated environment if unsure. 2) Confirm you have (or want) a Membrane account and understand that CLI login will store local tokens/credentials (check where membrane stores them). 3) Be aware the SKILL.md requires node/npm and network access even though the registry metadata doesn't declare them—if you cannot or will not install global npm packages, decline. 4) If you want to limit risk, disable autonomous invocation or run the CLI commands manually rather than permitting the agent to run them automatically.Like a lobster shell, security has layers — review code before you run it.
latestvk97ctcrgs1veaewzw0eh11pf09843tg7
License
MIT-0
Free to use, modify, and redistribute. No attribution required.