Hygraph

Security checks across malware telemetry and agentic risk

Overview

This Hygraph skill appears purpose-aligned, but it can change or delete live CMS content without clear confirmation safeguards.

Install only if you trust Membrane and want an agent to manage Hygraph. Use a least-privilege Hygraph/Membrane connection, verify the project and environment before each operation, and require explicit approval before publish, unpublish, update, delete, custom GraphQL, or raw proxy requests.

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: 87% confidence
Finding: The skill explicitly advertises destructive operations such as deleting assets and content entries, and publishing/unpublishing content, without requiring confirmation or warning about irreversible or production-facing effects. In an agent context, this increases the chance that a user request is interpreted too aggressively and leads to accidental data loss or unauthorized content state changes.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal