T08 · Insecure Dependencies
Warning
- Location
SKILL.md:127- Finding
Unpinned Package Execution Through npx
- Content
View full analysis
- Remediation
View remediation
Security audit
QUANT-X v10 量化策略仪表盘
Security checks for vulnerabilities and agentic risk
Overview
This is a live A-share stock dashboard, but it overstates its trading analytics and includes avoidable supply-chain risks users should review before relying on it.
Review before installing or publishing. Treat the trading scores and large-order panels as rough heuristics, not measured institutional/order-book data; avoid the unpinned `npx mdeploy` path unless you pin and verify the package; and consider self-hosting or integrity-protecting Chart.js before using the dashboard for decisions.
Vulnerability Patterns
- Insecure DependenciesIntroduces malicious components through unsafe dependency sources
- Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
- Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
- Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
- Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (4)
T08 · Insecure Dependencies
Warning
- Location
quant-x-v10.html:6- Finding
Remote JavaScript Dependency Loaded Without Integrity Protection
- Content
View full analysis
``` ### Technical Analysis The dashboard loads and executes JavaScript directly from a third-party CDN. Although the Chart.js version is fixed, the resource has no Subresource Integrity hash, and the page does not define a restrictive Content Security Policy. The browser therefore trusts whatever JavaScript the CDN returns for this URL. A compromise of the package distribution path, CDN infrastructure, or origin account could cause modified code to execute in every user's dashboard session. HTTPS protects transport confidentiality and integrity but does not protect against a compromised upstream source. ### Attack Path 1. An attacker compromises the relevant package-distribution or CDN delivery path. 2. The attacker causes modified JavaScript to be returned for the referenced Chart.js asset. 3. A user opens the dashboard. 4. The browser retrieves and executes the altered script because no integrity hash is present. 5. The script runs in the dashboard's page context. 6. It can manipulate displayed financial signals, initiate browser requests, inspect page-accessible data, or load further payloads, subject to browser-origin controls. ### Impact Assessment The attacker would obtain JavaScript execution in the dashboard page context. This permits complete manipulation of the dashboard interface and calculated trading signals. Depending on the hosting environment and browser state, the script could also access same-origin resources available to the page or induce users to act on falsified financial information. It does not directly grant operating-system privileges by itself. ]]>- Remediation
View remediation
``` The placeholder must not be deployed until replaced with the independently verified hash of the exact resource. ]]>
other
Warning
- Location
quant-x-v10.html:300- Finding
Synthetic Big-Order Metrics Presented as Measured Market Data
- Content
View full analysis
- Remediation
View remediation
T09 · Insecure Skill Coding Practices
Warning
- Location
quant-x-v10.html:235- Finding
Quote Fields Are Parsed Using Indices That Conflict With the Documented Schema
- Content
View full analysis
- Remediation
View remediation
Vulnerability Patterns
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (10)
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 96% confidence
- Finding
The skill documentation is entirely written in Chinese and does not state that the skill is China-specific or that users may choose another language. Under the policy criteria, forcing a specific language without opt-in or documented justification is a natural-language policy violation.
- Content
Rp1
Medium
- Category
- MCP Rug Pull
- Confidence
- 93% confidence
- Finding
The skill instructs users to run
npx mdeploy upload, which fetches and executes an npm package without pinning a specific version. If the package is updated maliciously, compromised upstream, or subject to dependency hijacking, a user executing this command could run attacker-controlled code on their system during deployment.- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 90% confidence
- Finding
The HTML root declares
lang="zh-CN", and the interface text is entirely Chinese, which effectively fixes the skill to a specific language/locale. The file does not offer localization options or explain that the skill is intentionally limited to a China-specific audience.- Content
Description-Behavior Mismatch
Medium
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The displayed OBI dimensions and big-order buckets imply access to granular order-book counts, time decay, and large-order segmentation. The implementation derives all OBI variants by scaling a single outer-vs-inner volume ratio and estimates big-order tiers as fixed percentages of total volume, which does not match the claimed depth of analysis.
- Content
Description-Behavior Mismatch
Medium
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
The dashboard presents named strategies and factor models as if they are analytically implemented. In code, these signals are derived from current-day percent change and amplitude thresholds only, without computing moving averages, Bollinger Bands, Donchian channels, 12-1 momentum, or Z-score mean reversion from suitable historical series.
- Content
Description-Behavior Mismatch
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The UI states "v10 整合 5 大量化框架 + Backtrader 回测", suggesting integrated backtesting functionality. However, the script only fetches live quotes, computes simple derived scores, and updates display widgets; there is no Backtrader usage, historical simulation engine, or replay/backtest logic anywhere in the file.
- Content
Intent-Code Divergence
Low
- Category
- Not specified by scanner
- Confidence
- 79% confidence
- Finding
The footer and header present a LIVE Tencent market dashboard, which implies an immediately interpretable local market timestamp. The tick function renders UTC time instead of China/local exchange time, creating an intent-level mismatch in the live-display semantics even though data fetching still occurs periodically.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 86% confidence
- Finding
The page automatically issues recurring requests every 3 seconds to an external market-data endpoint, causing user IP, user-agent, timing, and referrer metadata to be disclosed to a third party without a clear warning or consent flow. In this skill context, the behavior is expected for a live dashboard, which reduces suspicion, but the silent continuous exfiltration of browsing metadata still presents a privacy and network-governance risk.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 83% confidence
- Finding
This markdown file is entirely written in Chinese and does not indicate that language selection is optional or user-configurable. Under the stated policy, forcing a specific language without opt-in can be a natural-language policy violation unless the locale restriction is documented and justified.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
This markdown file contains user-facing operational instructions entirely in Chinese, and there is no indication that the user can choose another language or that the skill is intentionally limited to a Chinese-speaking audience. Under the natural-language policy rule, forcing a specific language without opt-in is a policy concern.
- Content