Submit Work

Security checks across malware telemetry and agentic risk

Overview

This skill is purpose-built for OpenAnt submissions, but it can automatically upload files and submit work externally without a final user confirmation.

Install only if you are comfortable with an agent sending task text and generated files to OpenAnt when it decides work is complete. Use explicit submission wording, review deliverable files before upload where possible, and avoid using it in workspaces that may generate sensitive or unrelated artifacts.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The skill description includes trigger phrases like 'I'm done', 'here's my work', 'deliver results', and 'upload and submit', which are broad enough to match ordinary conversation outside a clear OpenAnt submission context. That can cause the agent to invoke a side-effecting skill unexpectedly, leading to unintended task submission or file upload actions.

Missing User Warnings

High

Confidence: 97% confidence
Finding: The skill explicitly instructs the agent to 'always upload all output files without asking' and to submit immediately with 'No confirmation needed'. This creates a real exfiltration/privacy risk because locally generated files may contain sensitive data, credentials, internal notes, or unrelated artifacts, and the skill provides no user-consent or sensitivity check before external upload.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal