ClawHub
Back to skill

Security audit

competitive-intel-report

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to generate a static report while presenting it as data-driven competitive intelligence, so users should review it carefully before relying on it.

Review before installing or using for business decisions. Treat generated competitive, patent, infringement-risk, and market recommendations as a static template unless the agent separately performs and cites fresh PatSnap/Zhihuiya searches. Restrict the output path to a dedicated reports directory if running the script.

SkillSpector

Vulnerability Patterns
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (4)

Tainted flow: 'out' from os.environ.get (line 59, credential/environment) → pathlib.Path.write_text (file write)

Medium
Category
Data Flow
Content
out = Path(output_path)
out.parent.mkdir(parents=True, exist_ok=True)
out.write_text(html, encoding="utf-8")

print(f"✅ 报告已生成: {out}")
print(f"   公司: {our_company} ({our_short})")
Confidence
95% confidence
Finding
The script writes attacker-controlled content to an attacker-controlled path via environment variables without validation. If this skill runs in an automated agent or shared environment, a user can overwrite arbitrary writable files or place HTML output in sensitive locations, which is especially risky because the generated file may contain unescaped user-controlled HTML/JS content from the template substitutions.

Lp3

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding
The skill appears to use file read/write and environment-backed capabilities while declaring no permissions, which weakens transparency and can bypass normal user or platform expectations about what the skill is allowed to do. In a report-generation skill, undeclared filesystem access is concerning because it could read local templates, configuration, or other accessible files and write output without explicit permission disclosure.

Tp4

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding
The skill claims to generate data-driven competitive-intelligence analysis from company and patent inputs, but the finding indicates it only performs static template substitution without actually processing the patent results or computing the advertised analysis. This is dangerous because users may trust fabricated or stale output as if it were evidence-based intelligence, leading to bad business decisions and hidden integrity failures.

Description-Behavior Mismatch

Medium
Confidence
97% confidence
Finding
The template loads Chart.js from a third-party CDN, which contradicts the stated 'no external dependencies' property and creates a supply-chain and privacy risk. If the CDN resource is modified, blocked, or replaced in transit, the generated report executes untrusted JavaScript in the viewer's browser.

Static analysis

No suspicious patterns detected.