cn-funds-mcp

Security checks across static analysis, malware telemetry, and agentic risk

Overview

This appears to be a legitimate China fund/stock data helper, but it stores portfolio/reminder data locally and may proactively check reminders and offer generic investment suggestions.

Before installing, be comfortable with local storage of your fund holdings and reminders, and with queried fund/stock identifiers being sent to EastMoney-style market data APIs. Treat any add/reduce/stop-profit suggestions as generic information, not professional financial advice.

Static analysis

No static analysis findings were reported for this release.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Risk analysis

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

#
ASI01: Agent Goal Hijack
Low
What this means

The agent may automatically check reminders and potentially show due finance reminders before unrelated requests.

Why it was flagged

The skill explicitly instructs the agent to run a tool before responding in every conversation, even when the user asks about something else.

Skill content
每次对话开始时,必须先调用 `check_reminders` ... 无论用户说了什么,在回复用户之前,先调用 `check_reminders`
Recommendation

Install only if you want this proactive reminder behavior; remove reminders you no longer want and expect the first tool call in a conversation to be check_reminders.

#
ASI06: Memory and Context Poisoning
Low
What this means

Your fund codes, shares, and cost basis can remain on disk and affect later responses.

Why it was flagged

The skill stores personal fund holdings persistently and reuses them across sessions for profit calculations and reminders.

Skill content
持仓数据持久化在 `data/portfolio.json`,跨会话保留
Recommendation

Only save portfolio details you are comfortable storing locally; review or delete the data files if you no longer want the skill to remember them.

#
ASI09: Human-Agent Trust Exploitation
Low
What this means

The assistant may influence investment decisions even though it cannot trade for you.

Why it was flagged

The skill tells the agent to provide investment-style suggestions based on simple thresholds.

Skill content
对每只基金给出操作建议:... 可考虑适当加仓 ... 可考虑部分止盈 ... 建议评估是否继续持有
Recommendation

Treat these suggestions as informational only and verify with your own research or a qualified financial adviser before acting.

#
ASI04: Agentic Supply Chain Vulnerabilities
Info
What this means

Installation may depend on the current npm resolution rather than an exact locked dependency version.

Why it was flagged

The npm dependency uses a caret range, so a future install may resolve to a newer SDK version than the one originally tested.

Skill content
"dependencies": { "@modelcontextprotocol/sdk": "^1.12.1" }
Recommendation

Prefer installing from a trusted source with a lockfile or pin the dependency version if you need reproducible behavior.