Security audit

Root Cause Hunter

Security checks across malware telemetry and agentic risk

Overview

This debugging skill appears purpose-aligned, but it asks the agent to persist debugging notes and lessons without clear user control or scoping.

Review before installing if you do not want debugging details saved to persistent memory files. Use it only in workspaces where storing logs and lessons is acceptable, and avoid running it on issues that may expose secrets, customer data, credentials, or sensitive production details unless the skill is updated to ask before writing memory.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (4)

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill instructs the agent to record hypotheses, test steps, and results into `memory/debug_log.md`, which creates persistent workspace modifications without any explicit user consent or warning. In agent environments, silent writes can leak sensitive debugging data, overwrite existing notes, or create unexpected state that affects later runs.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: Directing the agent to write lessons learned into `memory/lessons.md` causes persistent modification of user or workspace data without warning, confirmation, or scope controls. This is risky because the file may not exist, may be shared across tasks, or may accumulate sensitive operational details that the user did not intend to store permanently.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The description embeds broad trigger phrases such as '诊断', '帮我调试', '帮我分析原因', and '排查问题', which are common user requests and can cause the skill to activate in many ordinary conversations. Unintended activation can redirect the agent into this skill's rigid workflow when the user did not explicitly request it, potentially disrupting normal behavior or overriding more appropriate handling.

Natural-Language Policy Violations

Medium

Confidence: 84% confidence
Finding: The skill description is written entirely in Chinese and specifies behavior in Chinese without indicating any language negotiation or user opt-in. This can lead to unintended language switching or inaccessible behavior for users operating in other languages, which is a policy and usability risk even if it is not directly code-execution related.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.