Back to skill

Security audit

社交轮播图设计 Social Carousel

Security checks for vulnerabilities and agentic risk

Overview

The skill supports user-approved carousel design and clearly discloses its dLazy login, credential storage, and cloud uploads.

Use this skill if you are comfortable sending your prompts and selected images to dLazy. Login saves an API key locally; an environment-variable option is documented. Confirm the generation plan and account costs before proceeding. The supplied static scan reported no findings.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied command configures authentication rather than implementing the declared carousel-design workflow. Nothing in the description or code establishes this as a supporting implementation detail. The chunk does not demonstrate credential exfiltration or use of a real API key.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied code invokes a login command rather than implementing the declared carousel-design workflow. No connection between this authentication operation and carousel design is established, so it cannot be treated as a demonstrated supporting implementation detail. The command alone does not establish how credentials are handled or whether login succeeds.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The supplied chunk only attempts to configure CLI authentication; it contains no carousel-design workflow, confirmation step, or cover-first execution. Nothing supplied establishes this authentication operation as a supporting requirement for the declared workflow. The command alone does not establish credential exfiltration or other network activity.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The supplied command performs a login operation rather than the declared carousel-design workflow. Nothing in the description establishes authentication through dlazy as a supporting requirement. The snippet does not reveal whether login succeeds or how credentials are handled.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.