agent-workflow-designer

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:17 PM UTC
• Version:
  1.0.0

Overview

This is a non-executable guide for designing multi-agent AI workflows; its external model API examples are expected for that purpose but require care with sensitive data.

Safe to install as a reference skill. Before adapting the code, use scoped API keys, avoid sending secrets, personal data, regulated data, or full internal documents in prompts and handoffs, redact unnecessary context, set spending and rate limits, and check your model provider's retention settings.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill repeatedly demonstrates sending user requests, intermediate outputs, competitor analyses, and workflow context directly to external model APIs, but it does not include any privacy notice, consent step, data classification guidance, or redaction requirements. In a production orchestration skill, this omission can lead users to forward sensitive business, personal, or regulated data to third-party providers without understanding the exposure, and the multi-agent design increases that risk by propagating data across several calls.

VirusTotal

65/65 vendors flagged this skill as clean.