RevOps Engine

• Outcome: Pass
• Latest audit: May 28, 2026, 4:39 AM
• Version:
  1.0.0

Overview

This is an instruction-only RevOps template skill with proportionate business guidance and no evidence of code execution, persistence, credential use, or data exfiltration.

Safe to install as a business-advisory skill. Use explicit prompts when invoking it, and avoid sharing confidential revenue, pipeline, customer, employee, or compensation data unless that is appropriate for your agent environment.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• 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

Findings (12)

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The trigger phrase "Audit our RevOps" is broad natural language that could be matched during ordinary conversation, causing the skill to activate when the user did not explicitly intend to invoke it. In an agent environment, ambiguous activation can redirect the model into this skill's instruction set and outputs unexpectedly, creating prompt-selection and scope-confusion risk even though the content itself is business-oriented.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: "Build our funnel definitions" is generic enough to overlap with normal planning dialogue, so the skill may be invoked accidentally during routine discussion about sales processes. That can cause unintended tool/skill selection and inject extensive prescriptive content into the conversation without clear user consent.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: "Create a pipeline review template" is a common business request and lacks any unique activation boundary, so ordinary user prompts may inadvertently activate the skill. This creates unintended behavior risk by allowing broad prompt overlap rather than deliberate invocation.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The phrase "Build our metrics dashboard" is highly generic and likely to occur in normal analytics discussions, making accidental invocation plausible. In multi-skill systems, this broad trigger increases the chance of incorrect routing to this skill and unexpected context takeover.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: "Design our lead scoring model" is a routine request in marketing and sales contexts and therefore not specific enough for safe skill activation. Because it can match ordinary user intent broadly, it raises the risk of unintentional invocation and prompt-scope confusion.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: "Create marketing-sales SLAs" lacks specificity and could easily be said in ordinary consulting or planning conversation, making accidental activation likely. The danger is not direct code execution but unintended skill engagement and instruction injection into the session flow.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: "Model our revenue plan" is broad and semantically common in finance and GTM discussions, so it may overlap with ordinary requests not intended to invoke this specific skill. This weak activation boundary can cause the wrong skill to run and shape the conversation unexpectedly.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The trigger "Score our RevOps maturity" is not sufficiently unique and could be naturally used in ordinary business dialogue, allowing unintended activation. In agentic systems, these broad commands are vulnerable to prompt-routing errors and hidden instruction selection.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: "Design our comp plan" is an especially common advisory request and is too generic to serve as a safe activation phrase. This can lead to accidental invocation during normal HR, finance, or sales compensation discussions.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: "Diagnose our funnel" is ambiguous and overlaps with ordinary growth or analytics conversation, so it does not provide a reliable activation boundary. The resulting risk is unintended skill invocation and context switching rather than direct system compromise.

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: "Build expansion signals" is generic enough to match normal RevOps or product-growth discussion and therefore may activate the skill unintentionally. In systems with automatic skill dispatch, that ambiguity materially increases the chance of misrouting and unrequested output generation.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: "Create our forecast model" is a common business request and lacks activation constraints, making it susceptible to accidental or overly broad matching. The skill context is operational rather than inherently dangerous, but ambiguous triggers still create real prompt-routing and unintended-execution risk.

VirusTotal

57/57 vendors flagged this skill as clean.