Security audit

Academic Deep Research

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only research workflow skill with disclosed web and memory use, and its heavier automated research phase is gated by user clarification and plan approval.

Install this if you want a structured, intensive research assistant. Prefer invoking it explicitly with /research, review the proposed plan before approving execution, narrow source counts or disable parallel sessions for sensitive topics, and treat it as an online research workflow despite the README's offline wording.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The README advertises broad natural-language triggers such as "deep research on..." and "exhaustive analysis of...", which are generic phrases likely to appear in normal conversation. In agent platforms that auto-route or invoke skills from user text, this can cause unintended activation, scope confusion, or prompt-collision with unrelated requests, especially because the skill is powerful and designed to perform extensive multi-step web research.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The trigger guidance includes broad phrases such as requests for exhaustive analysis or 'tell me everything about X,' which can cause the skill to activate during ordinary conversation rather than only when explicitly invoked. In this skill, unintended activation is more concerning because it can launch a long, tool-heavy research workflow using web_search, web_fetch, and sessions_spawn, increasing the chance of unnecessary external data access, cost, and user-surprising behavior.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The invocation guidance includes highly generic trigger text like "deep research" and "exhaustive analysis," which can be mentioned in ordinary conversation and unintentionally activate the skill. Because this skill performs extensive autonomous web activity in Phase 3, accidental invocation could lead to unexpected tool usage, wasted resources, or unintended external research on sensitive user topics.

Autonomous Decision Making

Medium

Category: Excessive Agency
Content: --- ## Phase 3: Research Cycles (Auto-Execute) ### Theme 1: Market Landscape — Cycle 1
Confidence: 92% confidence
Finding: Auto-Execute

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.