EasyClass Auto-Improve

• Outcome: Review
• Latest audit: May 30, 2026, 12:47 PM
• Version:
  1.2.0

Overview

This skill is a disclosed automation workflow, but it asks an agent to repeatedly change, commit, and push code without clear user approval boundaries.

Install only if you intentionally want a recurring autonomous coding workflow. Run it on a feature branch in a specific repository, protect main branches, require review before pushes or merges, and avoid putting credentials or sensitive instructions in TASKS.md.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Vague Triggers

Medium

• Confidence: 85% confidence
• Finding: The description markets the skill as broadly applicable for automated development on a recurring basis without defining explicit scope, approval gates, or execution boundaries. In a skill that can modify code and interact with git, ambiguous invocation language increases the chance of unintended autonomous execution on sensitive repositories or tasks the user did not mean to delegate.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The usage section explicitly frames the workflow as something that runs every 30 minutes, but it does not define stopping conditions, repository allowlists, approval checkpoints, or safe trigger boundaries. Because the workflow includes implementation, testing, committing, and pushing, an always-on loop can continuously make and publish unintended changes, amplifying mistakes or prompt-injection-driven behavior.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The documented workflow includes code changes, task updates, commits, and pushes, but it does not prominently warn users that the skill can perform automatic repository writes and remote publication. In practice, this can lead to accidental data loss, unsafe code landing in shared branches, or unauthorized publication of generated changes if users assume the skill is advisory rather than mutating.

VirusTotal

VirusTotal findings are pending for this skill version.