Security audit

Auto Updater Tool Free

Security checks across malware telemetry and agentic risk

Overview

This updater skill is mostly coherent, but it needs review because it can guide an agent to modify files, schedule checks, and run local callback scripts without tight safeguards.

Review this skill before installing. Use it only with trusted update sources and explicitly chosen target paths. Do not enable onUpdate or on-change scripts unless you wrote and reviewed the script, and check any cron entry it asks you to add. Use narrowly scoped tokens for private repositories and avoid exposing those values in logs or prompts.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (6)

Context-Inappropriate Capability

High

Confidence: 93% confidence
Finding: The skill documents command-execution hooks and positions itself as an updater, but the callback-style behavior allows execution of arbitrary local scripts during update-related flows. In an agent context with read/write/exec permissions, this expands the skill from version management into general code execution, which can be abused for persistence, destructive actions, or privilege misuse.

Context-Inappropriate Capability

High

Confidence: 96% confidence
Finding: The documented `onUpdate` / `on-change` callback behavior enables execution of arbitrary shell scripts after syncing or updating. Because the skill already advertises file modification and exec capabilities, a crafted configuration or prompt could use this hook to run unintended commands unrelated to safe update logic.

Intent-Code Divergence

Medium

Confidence: 81% confidence
Finding: The documentation is internally inconsistent: it states the free edition only notifies users, but elsewhere describes automatic script-driven update behavior. This can mislead users and reviewers about the actual execution and modification scope, increasing the chance that dangerous automation is enabled without informed consent.

Vague Triggers

Medium

Confidence: 78% confidence
Finding: The trigger conditions claim the skill should be used broadly for code generation, debugging, and deployment, which is much wider than a simple updater's stated purpose. Overbroad activation increases the chance the agent invokes a read/write/exec-capable skill in unrelated contexts, raising the risk of accidental or unsafe system changes.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill describes downloading files, modifying local state, backing up data, and running commands, but it does not prominently warn users that these actions can change the filesystem and execute system-level operations. In an agent environment, missing warnings reduce informed consent and make harmful actions more likely to be triggered automatically.

Missing User Warnings

Low

Confidence: 72% confidence
Finding: The skill instructs users to provide authentication through environment variables but gives no guidance on secret handling, redaction, rotation, or avoiding accidental disclosure in logs and subprocesses. While environment variables are common, omitting security guidance can still lead to credential leakage during agent-assisted execution.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.