Gitlab

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed GitLab integration using Membrane, with expected GitLab access and no evidence of hidden or malicious behavior.

Install this only if you trust Membrane and the npm CLI package. Review the GitLab permissions granted during connection setup, use the intended GitLab connection ID, and require confirmation before allowing the agent to create or update GitLab projects, issues, merge requests, or custom Membrane actions.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The skill documents create/update actions without explicitly warning that they modify live GitLab resources. In an agentic setting, this increases the chance an agent invokes a mutating action during exploration or ambiguous requests, causing unintended changes such as creating projects, issues, or merge requests.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal