sre-operator

• Outcome: Pass
• Latest audit: May 29, 2026, 8:15 PM
• Version:
  0.0.1

Overview

This is a server-operations helper with expected admin guidance and a read-oriented diagnostic script, but users should review high-impact commands before running them.

Install only if you want agent assistance with server operations. Treat every service, package, firewall, DNS, deletion, truncation, and privileged command as manual-review work: confirm the target host, take backups where relevant, prepare rollback steps, and redact sensitive logs before sharing them.

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 (5)

Vague Triggers

Medium

• Confidence: 79% confidence
• Finding: The trigger condition is very broad ('when users need system administration, performance optimization, log analysis, or server maintenance'), which can cause this skill to activate for a wide range of operational requests. In an agent environment, overbroad activation increases the chance that the skill will steer conversations toward shell-command execution or sensitive system guidance in contexts where a narrower or safer skill should apply.

Vague Triggers

Medium

• Confidence: 82% confidence
• Finding: The task goals repeat broad trigger conditions without defining boundaries, privilege assumptions, or prohibited actions. This makes the skill more dangerous because the content includes concrete administrative commands and references to scripts; if invoked too easily, it may encourage operationally risky actions on real systems without sufficient contextual checks.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: This section includes system-altering commands such as sending signals to processes and writing directly to /proc/sys/vm/drop_caches, but it does not consistently warn that these actions require elevated privileges and may disrupt workloads or affect performance. In an ops skill, users may copy-paste remediation commands during incidents, so insufficient guardrails can lead to unintended service interruption or unsafe execution.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: The guide recommends direct network and DNS changes such as iptables rule insertion, adding an IP address, and overwriting /etc/resolv.conf with shell redirection. These actions can immediately alter host connectivity or persistently break name resolution, and the document does not provide adequate warnings, rollback steps, or environment-specific caveats.

Missing User Warnings

High

• Confidence: 98% confidence
• Finding: This disk cleanup section contains irreversible commands including truncating logs, deleting files from /tmp, and cache cleanup actions that can remove forensic evidence or disrupt applications. Although one line says '先确认', the coverage is limited and does not adequately protect against accidental data loss when operators follow the guide under pressure.

VirusTotal

65/65 vendors flagged this skill as clean.