SettlementWitness

Security checks across malware telemetry and agentic risk

Overview

This skill appears to do what it says: verify task outputs through a remote verifier, with guidance to minimize submitted data.

Install only if you are comfortable sending minimized verification data to the configured verifier service. Do not submit secrets, private user data, proprietary outputs, or account/session details unless you have reviewed and trust the endpoint and its retention policy.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill directs users to submit task data to an external verifier service, which creates a real data exfiltration boundary. Although the document briefly says not to include secrets or private data, it does not clearly foreground that data is being transmitted to a third-party endpoint, what trust assumptions apply, or what privacy/security guarantees exist, so users may send sensitive task outputs or metadata off-platform.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal