ClawHub
Back to skill

Security audit

remember-projects

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward project-memory integration, but users should understand that project status, owners, milestones, and blockers are sent to an external BlueColumn/Supabase service.

Install only if you are comfortable sending project-management details to BlueColumn's service. Avoid including secrets, regulated data, sensitive HR details, or confidential blocker information unless your organization has approved that use, and prefer narrower recall queries when full project visibility is not needed.

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
92% confidence
Finding
The skill instructs users to send project descriptions, owners, milestones, blockers, and goals to an external Supabase-hosted endpoint without any privacy notice, data classification guidance, or limits on what may be transmitted. This can lead agents to exfiltrate sensitive business and personnel data to a third-party service by default, especially because the examples normalize sending detailed operational information.

Missing User Warnings

Low
Confidence
87% confidence
Finding
The recall workflow encourages fetching a broad summary of all active projects, owners, and blockers, but does not warn that the response may aggregate sensitive internal status, attribution, and risk information. In shared-agent or multi-user contexts, this increases the chance of overbroad disclosure to users or systems that do not need full visibility.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.