Security audit

Trade Journal AI Coach

Security checks across malware telemetry and agentic risk

Overview

This skill is a trade-journal coach, but it mixes retrospective analysis with strong future trading, sizing, and return guidance while asking users to upload sensitive financial records.

Review this skill carefully before installing. Use it only for retrospective journal analysis, redact account identifiers and unnecessary personal or financial details, and treat any sizing, setup-selection, instrument-focus, or expected-return language as non-authoritative educational output rather than trading advice.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Description-Behavior Mismatch

Medium

Confidence: 96% confidence
Finding: The skill says it does not provide future trade recommendations, but its example outputs include prescriptive directives such as concentrating on specific setups/instruments and forecasting expected next-week results. In a finance context, this contradiction can cause the agent to deliver de facto personalized trading advice despite the disclaimer, creating user-harm and compliance risk.

Intent-Code Divergence

Medium

Confidence: 95% confidence
Finding: The documentation presents the skill as 'not investment advice,' yet later examples directly instruct users to increase focus and sizing on certain setups and abandon others. That mismatch is dangerous because users may reasonably rely on the outputs as actionable trading guidance, especially when paired with authoritative metrics and psychology framing.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The trigger keywords are broad finance phrases like 'trading psychology' and 'review my trades,' which can cause the skill to activate in conversations that are only loosely related. Unintended invocation is risky here because the skill handles sensitive financial topics and may produce strong behavioral or trading guidance when the user did not explicitly request this workflow.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The invocation guidance is expansive and lacks clear boundaries for when the skill should not run, such as prospective trade planning, tax questions, or general market commentary. In this context, weak scoping increases the chance that the skill is applied to live-decision support rather than retrospective journaling, amplifying advisory and safety risks.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill actively solicits detailed trade logs, account balance, notes, and broker exports without warning about privacy, retention, or the sensitivity of financial data. That is dangerous because users may disclose highly sensitive personal and financial information, potentially including account identifiers or proprietary trading behavior, without informed consent or minimization.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.