Back to skill

Security audit

个人IP规划全流程助手

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly does what it claims, but it asks for unnecessary shell access and creates active HTML files with under-scoped write behavior.

Review before installing. The skill is not malicious from the inspected artifacts, but it should not need Bash, and users should be cautious about generated HTML files that load remote JavaScript and use names derived from user input. Prefer a version that removes Bash, sanitizes output filenames, confines writes to a report directory, and vendors or integrity-pins Chart.js.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (3)

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
SKILL.md:10
Finding

Unnecessary Unrestricted Shell Capability Violates Least Privilege

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:146
Finding

Generated Reports Execute Remote JavaScript Without Integrity Verification

Content
View full analysis
``` ### Technical Analysis The prescribed report template loads Chart.js directly from a third-party CDN whenever a generated report is opened. Although the dependency version is pinned, the template does not specify a Subresource Integrity hash and does not package a locally reviewed copy. Consequently, the effective JavaScript executed by a report is retrieved after the Skill itself has been reviewed. A compromise of the CDN, package publication channel, or dependency artifact could cause modified JavaScript to run in the report's browser context. Because the reports may contain business plans, competitor analysis, user profiles, and monetization strategies, malicious JavaScript could read or modify report contents and potentially transmit them over the network. The actual browser-level impact depends on how the report is opened and the browser's restrictions for local files. ### Attack Path 1. The Skill generates an HTML report containing the remote `
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:234
Finding

User-Controlled IP Name Is Used in an Output Filename Without Path Validation

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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
Findings (7)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger examples are natural-language phrases a user might say in ordinary conversation, and the README does not define strong activation boundaries. In an agent environment, overly broad triggers can cause unintended invocation of the skill, leading to surprise execution, inappropriate context capture, or interference with user intent routing.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest description lists many natural-language trigger phrases such as “个人品牌”, “打造IP”, “帮我规划个人IP”, and “IP怎么做”, which are broad and could match ordinary discussion rather than a clear intent to invoke this specific skill. The file does not provide exclusion conditions or negative examples to narrow when the skill should or should not activate.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill requests Bash even though its documented behavior only needs search, fetch, read, and write. Unnecessary shell access expands the attack surface: if later prompt content, fetched web data, or future edits influence tool selection, the agent may execute commands on the host without a legitimate need.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The instruction “使用纯中文,技术图表保留中英混排” imposes a specific language/locale on the generated report. There is no opt-in, language selection, or justification that this skill is limited to a Chinese-language regional context.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill mandates creating a local HTML file in the working directory without warning or obtaining explicit user consent. Silent file creation can surprise users, clutter shared environments, and increase risk when combined with generated active content such as HTML and external CDN scripts.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The natural-language user-facing description and trigger instructions are entirely in Chinese, which implies a fixed language experience without user opt-in. The file does not state that the skill is intentionally limited to Chinese-speaking users or provide an alternative language option.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The interaction flow says the skill should use open_result_view to display the generated report, but the manifest's declared tools list includes only WebSearch, WebFetch, Read, Write, and Bash. This is an intent/code-documentation mismatch because the documented execution steps rely on a capability that is not part of the declared implementation surface.

Content

No source excerpt is available for this finding.