Designers Eye

• Outcome: Pass
• Latest audit: Sep 12, 2026 · 2:08 AM UTC
• Version: 2.0.4

Overview

This is a markdown-only design critique skill that asks the agent to review user-provided visuals and does not request execution, persistence, credentials, or destructive authority.

Installers should understand that this skill will analyze visuals they provide and may comment critically on usability, accessibility, layout, color, and typography. Avoid sharing screenshots that contain private customer data, credentials, or confidential business information unless that content is intended for critique.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The invocation guidance says "Share any image. That's it," which is extremely broad and overlaps with many ordinary image-sharing contexts. The file does not provide clear trigger constraints, activation boundaries, or negative examples to distinguish when this skill should apply versus when an image is shared for some other purpose.

Autonomous Decision Making

Medium

• Category: Excessive Agency

Content

## 5. Error Prevention
Design to prevent errors before they happen.

**Violations:** Required fields with no indication, no inline validation, tiny touch targets, no confirmation before navigating away from unsaved work.

---

• 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

Users need undo/redo, escape routes, emergency exits. Let them get out of mistakes.

**What to look for:**
- Destructive actions without confirmation
- No back button or way to cancel an operation
- Modal with no close button (X or Escape key)
- Form with no clear way to reset/start over

• 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

Users need undo/redo, escape routes, emergency exits. Let them get out of mistakes.

**What to look for:**
- Destructive actions without confirmation
- No back button or way to cancel an operation
- Modal with no close button (X or Escape key)
- Form with no clear way to reset/start over

• 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

Users need undo/redo, escape routes, emergency exits. Let them get out of mistakes.

**What to look for:**
- Destructive actions without confirmation
- No back button or way to cancel an operation
- Modal with no close button (X or Escape key)
- Form with no clear way to reset/start over

• 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

- Form field marked required but easy to miss the indicator
- Credit card field accepting invalid formats silently
- Hover-only UI (not available on mobile/touch)
- No confirmation before navigating away from unsaved work

**Fix pattern:** Mark required fields clearly. Validate as users type. Confirm before danger.

• 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.

Natural-Language Policy Violations

Low

• Confidence: 90% confidence
• Finding: The guidance states that users scan in LTR languages using F- or Z-patterns, but the document presents this as general design advice without addressing right-to-left locales or clarifying that the recommendation is language-dependent. This can create a locale policy issue by implicitly centering one reading direction without user opt-in or explicit scope limitation.

Static analysis

No suspicious patterns detected.