Back to skill

Security audit

规划虾 Asset Planner

Security checks for vulnerabilities and agentic risk

Overview

This financial-planning skill is not malicious, but it asks agents to retain sensitive portfolio history and produce highly specific trading instructions with weak safeguards.

Install only if you are comfortable sharing sensitive financial screenshots and receiving model-generated allocation ideas. Do not let the agent store financial history unless you explicitly want that, remove account identifiers before upload, and independently verify any fund names, codes, amounts, fees, tax effects, and suitability before trading.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

other

Warning
Location
SKILL.md:112
Finding

Unbounded Retention of Sensitive Financial Information

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill instructions and user prompts are entirely in Chinese and present the assistant persona and workflow in that language only. There is no indication that users may choose another language or that the Chinese-only constraint is required for a documented regional-compliance reason.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly says to "remember everything," including historical portfolio allocations, risk preferences, and holding changes, which are highly sensitive financial data. Retaining this data without any retention limits, minimization rules, or user notice increases privacy risk and could expose a detailed financial profile if conversation memory is accessed, reused unexpectedly, or leaked.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The instruction to retain all historical portfolio and preference data indefinitely creates a persistent store of sensitive financial and behavioral information without justification or safeguards. In a financial-planning skill, long-term memory of balances, holdings, risk appetite, and transaction patterns can materially increase harm from unauthorized access, model misuse, or unintended cross-session disclosure.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The operational workflow includes a ready-to-execute trading checklist with exact products, platforms, and order amounts, but lacks a prominent warning at the point of action requiring user verification, suitability review, and acknowledgment of investment risk. In a financial-advice skill that analyzes personal holdings and proposes rebalancing, this makes the content more dangerous because users may treat generated instructions as authoritative and execute unsuitable trades without sufficient scrutiny.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The file's compliance section says the tool should avoid naming specific products as buy recommendations, yet the SOP and behavior rules elsewhere instruct the agent to output exact products, codes, and amounts. This creates a policy contradiction that can cause the skill to generate regulated, personalized investment instructions while presenting itself as a neutral calculation tool, increasing legal, trust, and user-harm risk.

Content

No source excerpt is available for this finding.