Back to skill

Security audit

FBoxCLI

Security checks for vulnerabilities and agentic risk

Overview

This looks like a real FBox CLI helper, but it needs Review because it combines high-impact industrial device control with an unpinned global CLI install and risky command-line credential examples.

Install only if you trust the FlexEM fboxcli distribution and understand that this skill can control or alter industrial IoT resources. Prefer a pinned and verified CLI version, avoid putting passwords or secrets directly in shell commands, use least-privilege FBox accounts, and require explicit operator review before writes, deletes, contact changes, location exports, alarm confirmations, or batch control operations.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T08 · Insecure Dependencies

Warning
Location
INSTALL.md:7
Finding

Unpinned Global Installation of an Externally Downloaded Binary

Content
View full analysis
Remediation
View remediation
``` 2. Publish SHA-256 checksums and cryptographic signatures for every platform binary. 3. Verify the binary signature or checksum before executing it. 4. Document the exact binary download origin and ensure downloads use an authenticated HTTPS endpoint. 5. Use reproducible builds and publish build provenance, such as npm provenance attestations or Sigstore signatures. 6. Avoid global installation where possible. Install into an isolated project directory or controlled execution environment with minimal privileges. 7. Lock dependency versions in the Skill's release process and periodically re-audit upgrades before changing the approved version. 8. Ensure the installation process fails closed if integrity or signature verification cannot be completed. ]]>

T09 · Insecure Skill Coding Practices

Warning
Location
INSTALL.md:37
Finding

Sensitive Credentials Passed Through Command-Line Arguments

Content
View full analysis
[OPTIONS] ``` ```bash fboxcli box add FB001234 fboxpassword --alias "Workshop 1" --group "Production Workshop" ``` ### Technical Analysis Secrets passed directly as command-line arguments can be exposed outside the target process. Depending on the operating system and environment, command arguments may be visible through process-listing tools, process telemetry, endpoint monitoring, audit logs, terminal recordings, crash reports, or command-history files. Shell history creates a particularly persistent disclosure risk because commands containing `--client-secret`, `-p`, or a positional device password may remain readable after the command finishes. Placeholder values in documentation reduce accidental disclosure in copied examples, but the documented workflow requires users to replace those placeholders with real secrets. `SKILL.md` instructs the Agent not to automatically populate usernames, passwords ...[truncated 1805 chars]
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (14)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The installation guide shows logging in with -u YOUR_USERNAME -p YOUR_PASSWORD directly on the command line, which exposes credentials to shell history, process listings, terminal logging, and possibly CI/job logs. In this skill context, the CLI manages industrial IoT devices, so compromised credentials could enable unauthorized access to device data or operational actions, making the example materially risky.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
77% confidence
Finding

The skill description advertises deleting devices, groups, monitoring points, contacts, and performing batch writes, all of which can affect user data or system integrity. The markdown does not provide any caution that these operations may be destructive or should be confirmed before use.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The README explicitly demonstrates password-based login with -u and -p password on the command line, which can expose credentials through shell history, process listings, terminal logging, and automation logs. In an industrial IoT administration context, leaked FBox credentials could grant access to device management, monitoring-point writes, alarm handling, and other sensitive operations.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The activation text includes broad triggers like “CLI 脚本”, “自动化运维”, and “批量操作设备”, which can match many general operational requests rather than only explicit FBox usage. In a high-impact industrial IoT context, over-activation increases the chance the agent invokes this skill unnecessarily and guides users toward device-management actions, including potentially destructive operations, in the wrong context.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The contact add command collects and writes personal information such as phone numbers and email addresses, but the documentation does not clearly warn that this is a state-changing operation involving PII. In an automation/CLI context, users may invoke it in scripts or bulk workflows without understanding privacy, compliance, or irreversible data-creation implications.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The contact update command modifies stored personal information, but the description does not warn users that it changes existing PII records. In an industrial operations skill where CLI commands may be automated, this increases the risk of accidental or unauthorized changes to notification contacts, potentially disrupting alerting and creating privacy/compliance issues.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
82% confidence
Finding

该技能文档标题与全文内容均仅使用中文呈现,没有说明这是面向特定中文环境的区域性文档,也未提供可选语言或语言适用范围说明。根据规则,强制特定语言且缺少用户选择或明确合理性说明,属于自然语言层面的语言/区域策略问题。

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documentation includes a concrete box add example with a plaintext device password directly on the command line. In an industrial IoT context, this is risky because users may copy the pattern into real operations, exposing credentials through shell history, process listings, logs, screenshots, or automation scripts.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill documents a command that retrieves device geolocation and street address data without any privacy, authorization, or data-handling warning. For industrial IoT deployments, location data can reveal facility whereabouts and operational footprint, which can aid physical targeting, reconnaissance, or unauthorized tracking if broadly exposed or casually shared.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

技能清单描述历史数据相关能力为“查询历史数据”,属于读取型操作;而该文档在 history delete 部分明确记录了删除历史记录条目的命令。这表示该技能在历史数据范围内不仅可查询,还可执行破坏性删除操作,超出了描述中对历史数据能力的表述。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

文件全文使用中文编写,但未说明该技能仅面向中文用户,也未提供其他语言选项或用户可选语言机制。若组织要求避免未经用户选择而强制特定语言,这属于自然语言层面的语言策略风险。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
72% confidence
Finding

The README is entirely written in Chinese and does not indicate that other languages are available or that Chinese is required for a region-specific reason. Under the stated policy, forcing a specific language without user opt-in can be a natural-language policy issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The instruction mandates converting displayed times to Beijing time (UTC+8) for users, which imposes a specific locale convention regardless of user preference or region. This is a natural-language policy concern because the skill does not offer a choice or document a justified region-specific limitation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

该技能文档全文使用中文,未见提供用户语言/区域选择,也未说明该语言限制的适用范围或业务理由。按规则,若技能强制特定语言而无用户 opt-in,属于自然语言层面的语言/locale 策略问题。

Content

No source excerpt is available for this finding.