ClawHub
Back to skill

Security audit

Auto Updater

Security checks across malware telemetry and agentic risk

Overview

This skill clearly does what it says, but it sets up unattended daily updates for Clawdbot and every installed skill, which is broad control over future agent behavior.

Install only if you are comfortable with Clawdbot and every installed skill updating automatically each day. Prefer using a dry-run or notification-only workflow, pinning or allowlisting trusted skills, and keeping a rollback path before enabling unattended updates.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • 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 (3)

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill explicitly encourages unattended updates of both the core bot and all installed skills on a daily cron schedule, but it does not warn users that this grants automatic trust to future upstream releases and can introduce breakage or malicious code without review. Because the action applies broadly to all installed skills and runs unattended, a compromised registry package, bad release, or incompatible update could be installed automatically and affect the user's environment before they notice.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The setup confirmation explicitly states that daily unattended updates will modify both the Clawdbot core and all installed skills, but it does not warn the user about the security and stability implications of automatic software changes. In the context of an auto-updater, this matters because future updates could introduce breaking behavior or malicious code through a compromised update source, and the user is not prompted to make an informed consent decision.

Self-Modification

High
Category
Rogue Agent
Content
# Capture new version
CLAWDBOT_VERSION_AFTER=$(clawdbot --version 2>/dev/null || echo "unknown")

# Update skills
log "Updating skills via ClawdHub..."
SKILL_OUTPUT=$(clawdhub update --all 2>&1) || true
echo "$SKILL_OUTPUT" >> "$LOG_FILE"
Confidence
94% confidence
Finding
Update skill

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.