Market Analyst

• Outcome: Pass
• Latest audit: Sep 12, 2026 · 7:30 AM UTC
• Version: 1.5.0

Overview

This is a broad financial market analysis skill with public market-data scripts; I found no hidden persistence, credential access, exfiltration, or trade execution behavior.

Install only if you want educational market-analysis assistance. It can make external market-data requests and may produce confident-looking buy, sell, allocation, stop-loss, and leveraged-trading suggestions, so verify current data, consider your risk tolerance and jurisdiction, and do not treat outputs as personalized financial advice or authorization to trade.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (44)

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 97% confidence

Finding

The description presents a comprehensive global investment advisory capability spanning many asset classes, both technical and fundamental analysis, and portfolio construction tasks. The supplied code is much narrower: it sends a POST request to TradingView's Vietnam scanner for one ticker on a VN exchange (default HOSE), computes technical summaries from returned fields (RSI, EMA, MACD, Bollinger Bands, ATR, stochastic, 52-week range), and prints a simple verdict. This is not merely a partial implementation detail; the primary purpose is materially narrower than declared. While the technical-indicator portion is consistent with part of the description, most advertised capabilities are unsupported by this code chunk.

Content

No source excerpt is available for this finding.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 97% confidence

Finding

The declared description presents a comprehensive investment analysis and advisory capability across many asset classes, including real-time technical indicators, fundamental metrics, portfolio optimization, market scanning, and broad advisory use cases. The supplied code does not implement those functions. It only performs one narrow subset of the description: DCA accumulation estimation. Even within that subset, it uses fixed hardcoded annual return assumptions and prints comparative projections; it does not fetch live data, analyze securities, optimize portfolios, or provide technical/fundamental research. Because the actual code’s primary purpose is much narrower and materially different from the declared broad analytical/advisory scope, this is a clear description-behavior mismatch.

Content

No source excerpt is available for this finding.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 97% confidence

Finding

The declared description presents a comprehensive multi-asset investment advisory and portfolio analytics skill. The supplied code is much narrower: it is a standalone CLI script for estimating historical returns and simple future scenarios for one VN stock ticker using VNDirect price history. While there is partial overlap with investment return estimation and use of Sharpe/CAGR, the code lacks most of the declared capabilities, including technical indicators, fundamental metrics, portfolio management, DCA planning, and support for non-VN assets. The primary purpose is therefore materially narrower and different from the declared description.

Content

No source excerpt is available for this finding.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 93% confidence

Finding

The code does partially align with the declared description in that it scans several global asset classes (US equities, crypto, forex, commodities) and uses TradingView-derived technical indicators including RSI, MACD, EMA, and Bollinger Bands. However, the declared purpose is much broader and presents the skill as a comprehensive investment analysis and advisory tool covering VN stocks, portfolio analysis, DCA planning, return estimation, optimization, and fundamental metrics like ROE, Sharpe, CAGR, and sector valuation. None of those broader capabilities are implemented in this code chunk. The code is narrowly focused on scanning a predefined set of markets and ranking instruments by a simple technical score. This is a material description-behavior mismatch due to overclaiming substantial unsupported capabilities.

Content

No source excerpt is available for this finding.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 95% confidence

Finding

The declared description presents a comprehensive cross-asset investment analysis and advisory tool with technical and fundamental analysis, portfolio construction, DCA planning, and return estimation across many asset classes. The supplied code only scans VN and US stock exchanges via TradingView endpoints, filtering by RSI, volume, and for US market cap, then scores results using technical signals such as EMA, MACD, and Bollinger Bands. It does not implement the claimed fundamental metrics, portfolio analytics, DCA calculations, or support for forex/crypto/commodities/ETFs in this chunk. Because the actual code’s primary purpose is much narrower and materially different from the declared broad advisory capability, this is a mismatch.

Content

No source excerpt is available for this finding.

Anti-Refusal Statement

High

• Category: Anti-Refusal
• Confidence: 82% confidence

Finding

The 'Always answer' instruction is an anti-refusal pattern that pressures the agent to respond even when data is missing, the request is high-risk, or the model should abstain. In a financial-advice skill, that can suppress important safety refusals and lead to overconfident or unsuitable recommendations.

Content

Scanner excerpt · SKILL.md (reported line 248)May include surrounding context.

md

- Infrastructure spending → industrials, materials
- GLP-1 drugs → LLY, NVO, healthcare sector

Always answer:
1. **Where is the sector in the cycle?** (growth / peak / recession / bottom)
2. **What are the catalysts for the next 1-3 years?**
3. **What are the main risks?**

Missing User Warnings

High

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

The forex and swing-trading sections provide leveraged execution details, entry/SL/TP structures, and position sizing without an equally strong upfront loss warning. Because leveraged products can produce rapid losses, this is especially dangerous in an agent skill that may deliver authoritative-seeming trade instructions to inexperienced users.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium

• Category: MCP Least Privilege
• Confidence: 84% confidence

Finding

The skill includes explicit network-capable commands and direct HTTP requests to TradingView endpoints, but it declares no tool scope or allowed-tools constraints. In an agent environment, this can enable unintended external access, make data egress harder to govern, and bypass least-privilege expectations for a financial-advice skill.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 93% confidence

Finding

The skill gives concrete buy/avoid recommendations early, before presenting strong warnings about investment risk and non-personalized guidance. In a finance context, this makes the content more dangerous because users may act on recommendations without seeing the limitations, which can cause material financial harm.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium

• Category: Excessive Agency
• Confidence: 75% confidence

Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 191)May include surrounding context.

Step 3: Valuation Analysis — MANDATORY (Stocks & ETF)

Do NOT recommend BUY without checking valuation.

P/E Analysis

text

Intent-Code Divergence

Medium

• Category: Not specified by scanner
• Confidence: 94% confidence

Finding

The skill presents strong buy/avoid signals, trade setups, and allocation guidance while disclaiming that it is 'for reference only' at the end. This contradiction can cause users or downstream systems to treat highly prescriptive financial outputs as safe informal guidance, increasing the risk of harmful reliance without proper suitability, risk, or compliance controls.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 94% confidence

Finding

This markdown file contains substantial instructional content in a mixed Vietnamese/English format, including core steps, examples, and warnings, but does not state any user opt-in or language-selection mechanism. That can violate a language/locale policy where skills should not force a specific language experience without explicit choice.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

This section gives concrete investment timing and allocation guidance such as when to increase, maintain, or stop DCA based on market indicators, but it does not include a clear warning that crypto investing is highly risky and that the content is not personalized financial advice. In a skill explicitly designed for asset selection and portfolio guidance, users may over-rely on these rules and suffer significant financial losses, especially in volatile crypto markets.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 94% confidence

Finding

The file names specific exchanges and gives a purchase flow for Vietnamese users (for example, P2P VND to USDT to BNB) without warning about exchange counterparty risk, custody loss, scams, stablecoin risk, or regulatory issues. Because this is operational guidance for acquiring crypto, users may treat it as endorsed safe procedure and expose funds to fraud, account freezes, or irreversible transfer mistakes.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 90% confidence

Finding

The document gives concrete portfolio allocations and purchase guidance without a clear disclaimer that outputs are educational only and may be unsuitable for a user's financial situation. In a skill explicitly designed to advise on what assets to buy, this increases the risk of harmful overreliance, especially for inexperienced users who may treat the recommendations as authoritative.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 93% confidence

Finding

These sections recommend high-risk assets and aggressive allocations, including crypto exposure, without adequately warning about volatility, drawdowns, liquidity risk, and total-loss scenarios. Because the skill's purpose is investment advisory across volatile asset classes, omission of safety framing can materially increase the chance that users follow speculative strategies that are inappropriate for their circumstances.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 92% confidence

Finding

The file turns unfolding geopolitical events into direct purchase instructions, such as prioritizing gold due to war risk, without emphasizing that such trades are speculative, timing-sensitive, and prone to whipsaw losses. Event-driven recommendations can prompt impulsive trading during stressed markets, making this particularly dangerous in a portfolio-advisory skill where users may act immediately on the guidance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 94% confidence

Finding

This markdown file contains substantial Vietnamese content alongside English headings and labels, which effectively forces a mixed-language presentation on all users. Under the policy rule, language constraints should either be justified as region-specific or offered as an explicit user choice.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 96% confidence

Finding

This section gives concrete investment timing, allocation, and product recommendations such as when to buy gold, target portfolio percentages, and a monthly DCA method, but it does so without a clear user-facing risk warning or suitability disclaimer. In a skill explicitly intended for investment advice, this increases the chance that users treat generic guidance as personalized financial advice, potentially leading to inappropriate risk-taking or losses.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 96% confidence

Finding

This markdown file is written primarily in Vietnamese but includes substantial English headings, labels, and instructions such as 'Key Signals', 'Portfolio Strategy', and 'Short-term'. The file does not indicate that the mixed-language format is optional, user-selected, or required for a region-specific purpose, which can violate language/locale policy expectations.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 89% confidence

Finding

This skill content includes substantial Vietnamese-only terms in tables, scenario labels, and the template output, which can effectively force a specific language/locale for users who may expect English guidance. The policy allows locale-specific content only when it is explicitly justified or when users are given a choice, neither of which is stated here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 88% confidence

Finding

The file is written primarily in Vietnamese and does not indicate that users can choose another language or that the content is intentionally limited to Vietnamese speakers. Under the policy, a skill that effectively forces a specific language without opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 93% confidence

Finding

This file gives direct investment recommendations, timing advice, and asset-specific buy/avoid guidance without any suitability caveat, risk disclosure, or prompt to consult a licensed financial advisor. In the context of a skill explicitly designed to provide portfolio and market advice, that omission increases the chance users treat the content as authoritative personalized financial guidance and make harmful decisions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

This markdown file contains substantial bilingual content, with many key instructions and investing notes written in Vietnamese while headings and labels remain partly in English. Because the skill does not state that it is intended specifically for Vietnamese-speaking users or provide an opt-in language choice, it can violate a language/locale policy requiring explicit user choice or justified locale constraints.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 97% confidence

Finding

The user-facing output strings are written in Vietnamese (for example 'Giá', '52 Weeks', and 'Conclusion' sections mixed with Vietnamese labels) with no option to select another language. This creates a natural-language locale policy issue because the script imposes a specific language on all users without opt-in or justification.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.