ClawHub
Back to skill

Security audit

UI Audit

Security checks for vulnerabilities and agentic risk

Overview

This is a UI/UX audit reference skill with mostly static markdown guidance, not a skill that takes hidden control of the agent or system.

This skill is reasonable to install for UI audit help. Review the install source as you would any npm or skill package, consider pinning versions for reproducibility, and treat privacy-related design suggestions such as geolocation defaults as prompts for careful product review rather than automatic implementation advice.

A.I.G

Vulnerability Patterns
  • 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
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (8)

Anti-Refusal Statement

High
Category
Anti-Refusal
Content
### Key Activities

☐ **Sketch 6+ low-fidelity concepts**
- Don't judge while creating
- Set a timer (15-30 min)
- Explore diverse approaches
Confidence
85% confidence
Finding
Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Direct Prompt Extraction

High
Category
System Prompt Leakage
Content
3. **Async validation** → Debounced (300-500ms)
4. **Cross-field validation** → On submit

### Error Display Guidelines

**DO:**
- Show error adjacent to the field
Confidence
85% confidence
Finding
Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.

Anti-Refusal Statement

High
Category
Anti-Refusal
Content
- Require customization to use the product
- Make defaults unusable
- Overwhelm with options
- Reset customization without warning
- Hide the reset/restore option

### Widget Patterns
Confidence
80% confidence
Finding
Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The README presents invocation examples like "Audit this design" and "Review this landing page for accessibility issues" without clarifying whether these are literal trigger phrases, examples only, or constrained to a particular skill-selection context. In a markdown skill description, this lack of trigger specificity can cause unintended activation because the examples resemble common everyday requests.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
- Destructive actions (delete, archive, remove)
- Significant changes (formatting, reorganizing)
- Accidental actions (easy to mis-click)
- Actions without confirmation dialog

### Implementation Patterns
Confidence
75% confidence
Finding
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
- Force users to complete tour
- Cover every feature
- Block user from exploring
- Repeat tour without consent
- Show tour to returning users

### Tour Step Anatomy
Confidence
75% confidence
Finding
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Missing User Warnings

Low
Confidence
84% confidence
Finding
The document recommends using geolocation for country defaults in forms but does not mention consent, transparency, fallback behavior, or jurisdictional privacy considerations. In a UI-audit skill, this can normalize privacy-invasive design guidance that downstream users may adopt without safeguards, leading to unnecessary collection or inference of location data.

Static analysis

No suspicious patterns detected.