ClawHub
Back to skill

Security audit

Context Guardian

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed local checkpointing helper for long-running agent tasks, with a real privacy consideration because it saves task state and summaries to disk.

Install this only where saving local task history is acceptable. Keep the configured context path inside the intended project, avoid putting secrets in goals or summaries, review retention/cleanup practices for .context_guardian, and consider stronger file permissions if the machine or workspace is shared.

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 (1)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
This code persistently writes task state, summaries, checkpoints, and event logs to disk under a project-controlled directory, but it provides no built-in notice, consent mechanism, retention control, or data minimization safeguards. In an agent context, these files can contain sensitive prompts, goals, file paths, decisions, and operational history; if stored on shared systems or in repos/workspaces with weak permissions, they can leak confidential information or create unintended durable records.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.