Security audit

市场盘前概览简报

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed A-share premarket market-briefing workflow, but it requires a daxiapi token and should be treated as informational financial analysis only.

Install only if you trust the daxiapi CLI/npm package and are comfortable configuring a daxiapi token locally. Avoid putting real tokens in shell history, screenshots, or synced dotfiles, and treat generated briefs as prior-close informational context rather than personalized trading advice.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (8)

Vague Triggers

Medium

Confidence: 83% confidence
Finding: The trigger phrases include broad terms like '今日市场' and '今日热点' that can match many generic market-related requests, causing the skill to activate outside its intended premarket scope. In an investing context, overbroad activation can lead to the agent invoking stale end-of-day data and presenting it as relevant to a user asking for broader or intraday analysis.

Vague Triggers

Medium

Confidence: 80% confidence
Finding: The additional trigger phrases remain ambiguous even though the document describes an 8:00-9:25 use window, because phrases like '盘前消息' and '今日市场' may still overlap with general financial commentary requests. This increases the chance of unintended skill routing, which is especially risky here because the skill produces structured investment-oriented output that could be mistaken for timely analysis despite relying on prior-close data.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The documentation shows `config set token YOUR_TOKEN` directly, which encourages users to place a secret on the command line. Secrets passed as CLI arguments are commonly exposed in shell history, process listings, terminal logs, or shared screenshots, creating a realistic credential leakage risk. In this skill context, the risk is somewhat elevated because the document is operational guidance and users are likely to copy-paste the command verbatim.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The document includes explicit investment-action language such as suggesting a possible buy opportunity when the fear/greed index is below a threshold, but it does not pair that guidance with a clear disclaimer that the content is informational only and not financial advice. In a premarket briefing skill intended to support investment decisions, users are especially likely to rely on these signals as trading recommendations, increasing the risk of harmful or unsuitable financial decisions.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: This section again presents a threshold-based 'buy signal' tied to trend temperature without any warning about investment risk, suitability, or the limitations of the metric. Because the skill is explicitly designed for premarket decision support, this framing can be interpreted as actionable trading advice rather than educational field documentation.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The phrase indicating a possible oversold rebound buying opportunity is explicitly action-oriented and may encourage speculative trading, yet no caution is provided regarding volatility, losses, or the non-advisory nature of the material. In the context of an A-share premarket briefing skill, this omission makes the content more dangerous because users may act on it immediately before market open.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The document explicitly instructs users to append an API token export to shell startup files like ~/.zshrc or ~/.bashrc, which stores the credential in plaintext on disk and may expose it to local compromise, backups, dotfile sync, or accidental sharing. Although common in developer docs, presenting this as a permanent setup path without a clear warning or safer alternatives increases the chance of credential leakage.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The guide tells users to run a command that prints the configured token value, which can reveal credentials in terminal scrollback, screen sharing sessions, logs, or recordings. In a token setup document, this is especially risky because users are likely to follow commands verbatim during onboarding.

Static analysis

No suspicious patterns detected.