Back to skill

Security audit

Aiwencai

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its financial-data purpose, but it needs review because it sends financial queries to an external API, uses an API key, forces unrelated promotional text, and ignores a cache-control option.

Review this skill before installing. It appears designed to query financial data, not to damage the system, but users should know that their query text is sent to an external iWencai API with an API key. Prefer setting IWENCAI_API_KEY as an environment variable instead of passing keys on the command line, avoid sending personal account details or confidential strategies, and be aware that the skill forces a promotional source phrase into responses.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T01 · Skill Instruction Hijacking

Warning
Location
SKILL.md:46
Finding

Mandatory Promotional Text Hijacks Agent Responses

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/cli.py:107
Finding

User-Selected Cache Control Is Silently Ignored

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding

The skill invokes code capabilities that use environment secrets and external network access, but it does not declare any tool scope or permission boundaries. This creates an authorization and transparency gap: a host system or reviewer cannot easily determine that the skill may access an API key and transmit user-derived queries to an external service.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill mandates that every response end with a fixed Chinese phrase, regardless of user preference or locale. While not directly enabling code execution or data theft, this is an instruction-integrity issue that can override user expectations, reduce transparency, and be used to impose hidden branding or policy content on all outputs.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documentation explicitly instructs passing the user's natural-language question directly to a third-party API, but it does not warn that prompts may contain personal, confidential, or regulated financial information that will be transmitted externally. In a finance-related skill, this increases privacy and compliance risk because users may ask account-specific or sensitive investment questions without understanding the data-sharing boundary.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The document explicitly recommends passing the API key via a command-line flag (for example, --api-key "your-key"). Command-line arguments are commonly exposed through shell history, process listings, audit logs, and CI/CD job logs, which can leak the bearer token to other local users or monitoring systems. In this skill context, the token grants access to a financial data API, so exposure can enable unauthorized API use and abuse of the account.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill instructs the agent to send rewritten user financial queries to an external API using an API key, but it does not warn users that their inputs will leave the local system. This is a privacy and consent issue, especially because financial queries can reveal sensitive investment interests or strategies.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
75% confidence
Finding

The file presents all instructions, parameter descriptions, and examples exclusively in Chinese, but does not state that the skill is China/Chinese-language specific or offer an alternative language option. Under the stated policy, forcing a specific language without user opt-in can be a natural-language policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

This code file contains user-facing natural-language strings such as the module docstring, argument descriptions, and usage examples exclusively in Chinese. Under the policy, forcing a specific language without offering the user a language or locale choice can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The argument parser and help text present --is-cache as a configurable request parameter, and query_basic_info accepts an is_cache argument. However, the payload hardcodes DEFAULT_IS_CACHE instead of using the provided is_cache value, so the documented behavior does not match the actual request sent.

Content

No source excerpt is available for this finding.