Security audit

accounting组合包

Security checks across malware telemetry and agentic risk

Overview

This finance bundle is not malicious, but it asks for broad file-writing, command-execution, API, and reminder-trigger authority around sensitive financial workflows without enough scoping or user-control guidance.

Review this before installing if you would use it on real accounting or business data. Use least-privilege credentials, keep backups, avoid in-place writes to source financial records, require confirmation before command execution or file mutation, and verify what APIs or member skills will receive sensitive data.

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 (5)

Missing User Warnings

Medium

Confidence: 82% confidence
Finding: The skill advertises read, write, and exec capabilities for finance workflows without a prominent warning that these operations can modify files and run commands. In a financial context, this increases risk because users may supply sensitive data and the bundle encourages end-to-end automation, making destructive or privacy-impacting actions easier to trigger without informed consent.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The example writes financial data back to a user-supplied path without warning about overwriting existing files, corrupting records, or altering regulated business data. Because the context is accounting/finance, accidental modification of source data can have outsized operational and compliance consequences.

Missing User Warnings

Medium

Confidence: 86% confidence
Finding: The installation guidance instructs users to configure API keys and call external interfaces without explaining secure credential storage, data sharing boundaries, or privacy implications. In a finance-related bundle, this omission is more dangerous because sensitive financial data and privileged credentials may be exposed through misconfiguration or unsafe handling.

Vague Triggers

Medium

Confidence: 83% confidence
Finding: The bundle's use-case text is extremely broad and effectively says each member skill should be used to complete related tasks and then combined, without defining activation criteria, task boundaries, or safety constraints. In a bundle that exposes read, write, and exec across multiple finance-oriented skills, this ambiguity can lead to overbroad invocation, unexpected chaining, and unsafe handling of sensitive financial workflows.

Missing User Warnings

High

Confidence: 94% confidence
Finding: The manifest advertises write and exec capabilities for all bundled skills but provides no warning that these actions may modify files, trigger commands, or otherwise affect the host environment. Because the bundle is positioned for end-to-end finance processing, users may expose sensitive records or permit impactful operations without understanding the operational risk.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.