Html Coder Paid

• Outcome: Review
• Latest audit: Aug 2, 2026 · 9:31 AM UTC
• Version: 1.0.0

Overview

This HTML helper is not malicious, but it asks for broad read, write, command, and external-integration authority without tight boundaries.

Install only if you are comfortable with a skill that may read and modify project files, search repositories, run commands, and potentially use external APIs. Keep it limited to explicit HTML or web-audit tasks, review commands before execution, and avoid giving it credentials or broad repositories unless the workflow truly requires that access.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• 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 (6)

Description-Behavior Mismatch

Medium

• Confidence: 88% confidence
• Finding: The skill is presented as an HTML coding tool but claims additional security auditing, dependency vulnerability detection, and CI/CD review capabilities that are outside its stated scope. This kind of scope expansion can mislead users into granting broader trust and invoking higher-risk workflows, especially because the skill also declares exec/write capabilities.

Description-Behavior Mismatch

Medium

• Confidence: 85% confidence
• Finding: The declared output format is an audit/scoring report rather than HTML generation, which is inconsistent with the advertised purpose of an HTML coding tool. That mismatch increases the chance that the skill will be used as a broad analyzer or reviewer and may justify unexpected file inspection, command execution, or reporting behavior beyond what users anticipated.

Context-Inappropriate Capability

Medium

• Confidence: 83% confidence
• Finding: The documentation advertises generic external API integration even though that capability is not necessary for a local HTML coding helper. Unjustified external integration expands the data exfiltration surface and may allow project content, prompts, or metadata to be sent to third-party services without a clear need tied to the skill's purpose.

Intent-Code Divergence

High

• Confidence: 94% confidence
• Finding: The skill states that it is pure Markdown and needs no extra configuration, but later discusses API authentication failures and external API use. This inconsistency is dangerous because it obscures the true trust boundary and can cause users to believe no credentials or outbound access are involved when in fact the skill may depend on them.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The skill advertises read, write, exec, glob, and grep capabilities near the top-level metadata without a clear upfront warning about system-impacting behavior. In the context of an HTML tool, those permissions are broader than necessary and could enable unauthorized file modification, environment discovery, or command execution if the skill is invoked too broadly.

Vague Triggers

Medium

• Confidence: 82% confidence
• Finding: The invocation instructions are generic and do not define a narrow trigger scope, so the skill could be applied to a wide range of requests without clear boundaries. Combined with powerful tools like exec and write, broad activation criteria increase the risk of accidental overreach into unrelated files, commands, or workflows.

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

No suspicious patterns detected.