Post Update Awareness

Security checks across malware telemetry and agentic risk

Overview

This skill transparently checks OpenClaw update details and reports them without performing updates, installs, or configuration changes.

Install this if you want unattended OpenClaw update summaries. Be aware it can run local OpenClaw diagnostic commands, inspect plugin/channel status and update-guard logs, make an unauthenticated GitHub request for release notes, and write a small local state file. The reviewed artifacts do not show automatic updates, config edits, dependency installs, credential handling, telemetry, or destructive behavior.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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
Findings (1)

Intent-Code Divergence

High
Confidence
95% confidence
Finding
The skill claims to be read-only and 'does not mutate state,' yet it persistently writes a state file and retry counters. That contradiction matters because users may install or auto-run the skill under a false assumption that it performs no mutations, undermining trust boundaries and potentially bypassing policies that allow read-only skills but restrict persistent writes.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal