Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md → Step 1: 调用 scripts/health_check.cjs 执行扫描
Security audit
Pc Health Check
Security checks for vulnerabilities and agentic risk
Overview
This skill performs the Windows health-check behavior it advertises and does not show evidence of hidden exfiltration, persistence, or destructive activity.
Install only if you are comfortable with a local Windows diagnostic collecting system, process, startup, event log, listening port, and update details. Avoid sharing generated reports without reviewing or redacting host-specific information first.
Vulnerability Patterns
- Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
- Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
- Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
- Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
- Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- 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
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (11)
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md → Step 1: 调用 scripts/health_check.cjs 执行扫描
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The README advertises collection of system, process, network, port, driver, event, startup, and update information without warning that the generated report may contain sensitive local metadata. Users may invoke or share the report without understanding that it can expose host details useful for fingerprinting, internal network mapping, or privacy leakage.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 84% confidence
- Finding
This README is a markdown file, so trigger quality applies. The phrase "电脑健康检查" is generic enough that it could overlap with ordinary requests about computer health or troubleshooting, and the document does not provide boundaries, negative examples, or a more explicit activation context to reduce unintended invocation.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 96% confidence
- Finding
The description does not warn users that the skill may collect sensitive system and network details, including processes, listening ports, drivers, event logs, and security update state. Without upfront disclosure, users may invoke the tool without understanding the privacy implications, increasing the risk of unintended exposure of internal system configuration or security posture.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The trigger phrases are broad enough to match routine troubleshooting requests such as checking system status, which can cause the skill to activate when the user did not explicitly consent to a full health scan. Because this skill collects detailed host, network, process, port, driver, and update information, accidental invocation can expose sensitive environment data and initiate system inspection beyond user intent.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
The code explicitly formats timestamps with the zh-CN locale and emits the generated report text in Chinese, with no option for the user to choose language or locale. This is a natural-language policy concern because it imposes a specific language/locale rather than offering an opt-in or configurable choice.
- Content
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 91% confidence
- Finding
The manifest says the tool generates a structured Markdown report with risk levels and handling advice across health checks. However, these sections are only rendered when status is 'ok', so warning/error results for CPU, memory, and disk are omitted from the Markdown report instead of being highlighted.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 89% confidence
- Finding
This code performs a filesystem write using a user-supplied path, which can modify local data, but the only disclosure is a post-write success message. There is no confirmation prompt, pre-write warning, or explanatory comment/docstring near the save behavior to alert users before data is written.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 87% confidence
- Finding
Natural-language policy checks apply to all file types. The file forces a specific language for user-facing instructions and invocation examples, with no opt-in, alternative language, or explanation that the skill is intentionally region- or audience-specific.
- Content
Description-Behavior Mismatch
Low
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The manifest explicitly states that v1.0.1 added support for --save, --silent, and --version. While the code does implement those flags, the exported version string remains '1.0.0', so the skill's self-reported version does not match the documented release state.
- Content