Alibaba Cloud
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This looks like a real Alibaba Cloud integration, but it gives an agent broad cloud-management power through Membrane credentials without clear limits or confirmation rules.
Install only if you trust Membrane and the npm CLI package. Use a least-privileged Alibaba Cloud RAM user or role, verify every high-impact action before it runs, avoid broad production/admin access unless required, and revoke the Membrane connection when you are done.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
A mistaken or overly broad agent action could alter cloud infrastructure, security policy, secrets, or production workloads.
The skill exposes generic action use over high-impact Alibaba Cloud resources, including security controls, Kubernetes secrets, orchestration stacks, and autoscaling, but does not show clear confirmation or containment requirements for risky changes.
- **Cloud Firewall Control Policy** - **Cloud Firewall Vpc Firewall** - **Container Service Kubernetes Cluster** - **Container Service Kubernetes Node Pool** - **Container Service Kubernetes Application** - **Container Service Kubernetes Namespace** - **Container Service Kubernetes Secret** - **Container Service Kubernetes Service** - **Resource Orchestration Service Stack** - **Auto Scaling Group** - **Domain** - **Ddos Protection Plan** - **Waf Instance** - **Actiontrail Trail** Use action names and parameters as needed.
Use least-privileged Alibaba Cloud credentials, require explicit user confirmation for create/update/delete or security-impacting actions, and prefer read-only discovery before mutation.
If broad Alibaba Cloud credentials are connected, the agent and Membrane-backed actions may be able to access or change sensitive cloud resources.
The skill delegates Alibaba Cloud authentication and ongoing credential refresh to Membrane, but does not document required scopes, least-privilege setup, or what level of Alibaba Cloud authority should be granted.
Membrane handles authentication and credentials refresh automatically ... "connect" — user needs to authenticate (OAuth, API key, etc.).
Connect only a restricted RAM user or role scoped to the specific task, avoid production/admin credentials unless necessary, and revoke the Membrane connection when finished.
The behavior depends on the currently published npm CLI version, which may change over time.
The skill relies on a globally installed external CLI from npm using the moving @latest tag. This is disclosed and purpose-aligned, but users should notice the unpinned package and global install.
npm install -g @membranehq/cli@latest
Verify the package source, consider pinning a known CLI version, and install in an isolated environment where possible.
External setup instructions could influence how the agent proceeds during connection setup.
The Membrane service may return dynamic instructions for the agent. This is part of the stated workflow, but such instructions should be treated as tool output rather than allowed to override the user’s intent or safety checks.
`clientAction.agentInstructions` (optional) — instructions for the AI agent on how to proceed programmatically.
Follow only instructions that are consistent with the user’s request, and do not let provider-returned instructions bypass confirmation for sensitive cloud actions.