ClawHub

Cursor Council

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent multi-agent Cursor orchestration guide, but it encourages forceful and lightly reviewed agent actions that can change a workspace and retain sensitive prompts or outputs.

Install only if you are comfortable supervising multiple agents that may edit code or run commands with reduced approval prompts. Use disposable workspaces or separate git branches, inspect approvals instead of blindly sending `y`, avoid secrets or customer data in council prompts, and delete or protect saved council transcripts.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (2)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The guide instructs users to paste a framed question into external model prompts and then capture all model outputs into a local markdown file, but it provides no warning about sending potentially sensitive architecture, security, business, or project data to third-party services and persisting that data on disk. In a skill explicitly used for architecture decisions and risk assessment, users are likely to include confidential internal context, making inadvertent data disclosure a realistic risk.

Ssd 3

Medium
Confidence
94% confidence
Finding
The skill explicitly instructs users to persistently archive full prompts and full model outputs, which can easily contain secrets, proprietary code, internal architecture details, credentials pasted into prompts, or regulated data. Because the archive path is long-lived and centrally organized, this increases the chance of sensitive data retention, secondary disclosure, and accidental inclusion in backups, syncing tools, or source control.

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal