AI Product Radar

• Outcome: Review
• Latest audit: Aug 5, 2026 · 2:13 PM UTC
• Version: 1.0.1

Overview

The skill largely matches its stated product-monitoring purpose, but its default screenshot path can execute generated code influenced by feed-provided links, so it needs review before use.

Review this skill before installing or running it. Use `--no-screenshots` unless the screenshot code is fixed to pass URLs and paths as arguments or JSON-escaped values, and run it only in an environment where outbound browsing to RSS feeds and linked product pages is acceptable.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Output Handling: Unvalidated Output Injection, Cross-Context Output, Unbounded Output
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Behavioral AST: exec() Call, eval() Call, Dynamic Import
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration

Findings (5)

subprocess module call

Medium

• Category: Dangerous Code Execution

Content

await browser.close();
        }})();
        """
        result = subprocess.run(["node", "-e", script], capture_output=True, text=True, timeout=45)
        return result.returncode == 0 and path.exists()
    except Exception:
        return False

• Confidence: 98% confidence
• Finding: result = subprocess.run(["node", "-e", script], capture_output=True, text=True, timeout=45)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 90% confidence
• Finding: The skill advertises and documents behavior that performs network access, file reads/writes, and shell execution via a Python script and cron usage, but no explicit permissions are declared. This creates a transparency and governance gap: users or calling systems may invoke a capability-rich skill without clear consent boundaries, increasing the risk of unintended external requests, filesystem modification, or command execution in sensitive environments.

Vague Triggers

Medium

• Confidence: 82% confidence
• Finding: The trigger list includes broad natural-language phrases such as 'or similar product intelligence workflows,' which can cause the skill to activate in contexts the user did not clearly intend. Over-broad invocation is dangerous here because the skill performs external network fetching and screenshot capture, so accidental activation can lead to unintended data collection, outbound requests, and local file creation.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill description explains outputs and pipeline stages but does not clearly warn users that it will contact external RSS feeds and product websites and may capture screenshots of those sites. This omission reduces informed consent and can expose users to privacy, compliance, and operational risks, especially in restricted environments where outbound browsing or storing fetched content is sensitive.

Unvalidated Output Injection

High

• Category: Output Handling

Content

await browser.close();
        }})();
        """
        result = subprocess.run(["node", "-e", script], capture_output=True, text=True, timeout=45)
        return result.returncode == 0 and path.exists()
    except Exception:
        return False

• Confidence: 97% confidence
• Finding: subprocess.run(["node", "-e", script], capture_output

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

No suspicious patterns detected.