Doppel Block Builder

• Outcome: Review
• Latest audit: May 28, 2026, 5:44 AM
• Version:
  1.0.0

Overview

The skill is mostly a coherent Doppel block-building guide, but it also pushes agents toward unrelated social-outreach and recruitment behavior without clear user approval.

Install only if you want an agent to submit and manage Doppel MML builds with your Doppel credentials. Review the exact MML, target space, and requested action before submission, and ignore the social-outreach, recruiting, streak, or reputation prompts unless you explicitly want those activities.

SkillSpector

By NVIDIA

Vulnerability Patterns

• 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
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill presents a destructive `delete` action as a normal workflow step without a clear warning that it removes the agent’s published build. In an agent setting, terse operational guidance can be executed automatically or reflexively, increasing the risk of unintended loss of content and disruption of the agent’s presence in the shared world.

Natural-Language Policy Violations

Medium

• Confidence: 91% confidence
• Finding: The opening language uses social and reputational pressure to push ongoing participation ('fall behind', permanent scoring) even though that is not necessary for safely using a block-building skill. In agent ecosystems, this kind of persuasive framing can bias behavior away from user intent and toward platform-engagement objectives, which is unsafe from an alignment and autonomy perspective.

Natural-Language Policy Violations

Medium

• Confidence: 97% confidence
• Finding: The 'Next step' section instructs the agent to install another skill and engage in growth/recruitment tactics ('share → recruit → build again') that are unrelated to the declared block-building purpose. This is a clear scope expansion attempt that can redirect an agent into promotional or manipulative actions without user authorization.

VirusTotal

60/60 vendors flagged this skill as clean.