Deep Research

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed deep-research helper that saves local research notes in a workspace ledger, with no evidence of hidden network, credential, or destructive behavior.

Install only if you are comfortable with research notes, source URLs, and claim summaries being saved in the workspace under .deepresearch. Use simple slugs without path separators, and review saved ledgers before sharing a project because they may contain sensitive research context.

SkillSpector

By NVIDIA

Vulnerability Patterns

MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
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 (1)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 95% confidence
Finding: The skill instructs the agent to read from and write to the workspace by creating and updating `.deepresearch/<slug>/` state and invoking `scripts/deepresearch.py`, but it does not declare any corresponding permissions. This creates a capability/permission mismatch that can bypass operator expectations, making persistent file modification possible without explicit review; in this context, the resumable ledger design increases exposure because repeated writes are a core workflow feature.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal