Daily Game News

• Outcome: Pass
• Latest audit: May 29, 2026 · 8:15 PM UTC
• Version: 2.0.0

Overview

This skill is a disclosed game-news crawler and report generator, with some privacy and transport-security caveats but no artifact-backed malicious behavior.

Install this only if you are comfortable with a scheduled crawler making outbound requests, storing reports locally, and potentially sending report content through your messaging workflow. Avoid using it for internal or sensitive URLs unless you add URL allowlisting, disable third-party proxy fetching, and restore normal TLS verification in the RSS helper.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Behavioral AST: exec() Call, eval() Call, Dynamic Import
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (8)

subprocess module call

Medium

• Category: Dangerous Code Execution

Content

try:
        # 使用 curl 抓取 HTML
        result = subprocess.run(
            ['curl', '-s', '-A', 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36', 
             '--max-time', '30', '-L', target_url],
            capture_output=True, text=True, timeout=35

• Confidence: 80% confidence
• Finding: result = subprocess.run( ['curl', '-s', '-A', 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36', '--max-time', '30', '-L', target_url], captu

subprocess module call

Medium

• Category: Dangerous Code Execution

Content

script_path = os.path.join(os.path.dirname(__file__), 'web_fetch_wrapper.py')
        
        if os.path.exists(script_path):
            result = subprocess.run(
                ['uv', 'run', script_path, url],
                capture_output=True, text=True, timeout=30
            )

• Confidence: 84% confidence
• Finding: result = subprocess.run( ['uv', 'run', script_path, url], capture_output=True, text=True, timeout=30 )

subprocess module call

Medium

• Category: Dangerous Code Execution

Content

def fetch_html(url, timeout=30):
    """使用 curl 抓取 HTML"""
    try:
        result = subprocess.run(
            ['curl', '-s', '-A', 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36',
             '--max-time', str(timeout), '-L', '--compressed', url],
            capture_output=True, text=True, timeout=timeout + 5

• Confidence: 88% confidence
• Finding: result = subprocess.run( ['curl', '-s', '-A', 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36', '--max-time', str(timeout), '-L', '--compressed', url],

Intent-Code Divergence

Medium

• Confidence: 89% confidence
• Finding: The module claims it performs direct crawling, but for some sites it silently sends requested URLs through the external r.jina.ai proxy service. This creates an unexpected data-flow boundary: requested targets and fetched content are disclosed to a third party, which can violate operator expectations, privacy assumptions, or compliance requirements in a security-sensitive environment.

Intent-Code Divergence

Medium

• Confidence: 98% confidence
• Finding: The script explicitly disables TLS certificate verification for both the requests and urllib code paths, which allows a man-in-the-middle attacker to intercept or modify RSS content fetched over HTTPS. Because the fetched feed data is then parsed and presented as trusted output, users may consume attacker-controlled content without any indication that transport security was bypassed.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The README states that reports are pushed via Feishu private message and stored locally as Word documents, but it does not clearly warn users that fetched third-party content and metadata will be transmitted off-process and retained on disk and in logs. This creates a transparency and privacy issue: users may invoke the skill without understanding where data goes or how long it persists.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The skill explicitly states that it will send reports via Feishu and archive Word/text copies to fixed local paths, but the documentation does not present this as a user-facing warning or consent point. This can lead to unintended disclosure or retention of potentially sensitive browsing, reporting, or business intelligence data, especially in an automated daily workflow.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: When use_web_fetch is enabled, the code sends the destination URL to https://r.jina.ai/, which discloses browsing targets and causes a third party to retrieve content on the user's behalf. This creates a data-leak/privacy issue and changes the trust boundary without clear user consent, which is especially risky in an agent skill that may be run against sensitive or internal targets from configuration.

VirusTotal

66/66 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.