Back to skill

Security audit

vibe-interaction-spec

Security checks for vulnerabilities and agentic risk

Overview

This frontend design skill uses proportionate local tools, with a minor privacy caveat in its advice about saving form drafts.

Run the linter against the intended project paths. If you implement draft saving, make it visible and optional, exclude sensitive fields, and provide expiration and deletion controls. The skill otherwise requires no credentials or elevated access.

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
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_read.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Ae4

Medium
Category
analysis-evasion
Confidence
80% confidence
Finding

Suspicious Unicode normalization or mixed-script content

Content

No source excerpt is available for this finding.

The wizard component prescribes saving form drafts locally without warning users that their entered data will persist.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

The wizard specification recommends persisting form drafts locally without limiting which fields are saved, defining expiration, or notifying users before persistence. If applied to sensitive forms, this can leave entered information accessible to subsequent users of a shared browser profile or to same-origin scripts; the general-purpose component context makes unsafe reuse plausible, although no actual sensitive-data storage implementation is shown.

Content

No source excerpt is available for this finding.

Terminology output is required to include Chinese and English without user opt-in.

Low
Category
Output Handling
Confidence
96% confidence
Finding

The instruction requires every term to have a Chinese name plus an English name, regardless of the user's preferred language. Although it explains the usefulness of English implementation terminology, it does not justify requiring Chinese or offer a language choice.

Content

No source excerpt is available for this finding.