Security audit

lita-sales-coach

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only Chinese sales-coaching skill with broad coaching triggers but no hidden execution, data access, or persistence beyond installation.

Install this only if you want a Chinese-first sales communication coach. Use explicit prompts when invoking it, and for financial or regulated sales scripts, treat its output as draft wording that still needs review against your institution's compliance rules and the customer's actual suitability.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (5)

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The README describes invocation in very broad, open-ended terms, including many sales, persuasion, reporting, and communication-coaching scenarios without firm scope boundaries. In agent environments, overly broad trigger criteria can cause unintended activation on unrelated conversations, leading the skill to influence sensitive user interactions or override more appropriate domain-specific skills.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The trigger text is extremely broad and includes generic communication, persuasion, negotiation, reporting, resilience, and 'help the other party make a suitable choice' scenarios. That can cause the skill to activate outside a narrow sales-coaching context and inappropriately steer unrelated conversations toward persuasive sales framing, especially in sensitive financial contexts.

Natural-Language Policy Violations

Medium

Confidence: 72% confidence
Finding: The skill strongly enforces a specific Chinese sales-coaching style and response format without indicating user override, which can reduce user agency and produce mismatched outputs for users expecting another language, tone, or compliance framing. In isolation this is not a classic security flaw, but it can become a safety and quality issue when applied in regulated or cross-lingual contexts where precise user-directed phrasing matters.

Natural-Language Policy Violations

Medium

Confidence: 90% confidence
Finding: This module is written entirely in Chinese and prescribes fixed Chinese output templates without indicating any ability to adapt to the user's preferred language or locale. In a sales-coaching skill, especially one covering finance and compliance-sensitive scenarios, forcing a single language can cause users to misunderstand scoring, compliance caveats, and remediation guidance, which increases the risk of miscommunication and inappropriate use.

Natural-Language Policy Violations

Medium

Confidence: 95% confidence
Finding: This file is entirely written in Chinese and presents no indication that the skill will detect or honor the user's language preference. That can cause incorrect or inaccessible responses for users expecting another language, which is especially problematic in financial and sales coaching contexts where misunderstanding wording, compliance guidance, or negotiation framing can materially affect outcomes.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.