语音转文字 ElevenLabs STT

• Outcome: Pass
• Latest audit: Sep 20, 2026 · 2:31 AM UTC
• Version: 1.3.18

Overview

This skill is a disclosed dLazy CLI wrapper for cloud speech-to-text, with some documentation inconsistencies but no artifact-backed malicious behavior.

Before installing, understand that audio files and parameters you provide will be sent to dLazy-hosted services, not processed only locally. Use `npx @dlazy/cli@1.2.3` if you do not want a global CLI install, avoid uploading sensitive audio unless dLazy's terms fit your needs, and rotate or revoke the dLazy API key if needed.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (7)

Ae1

High

• Category: analysis-evasion

Content

[English](./SKILL.md) · [中文](./SKILL-cn.md)

• Confidence: 100% confidence
• Finding: Referenced artifact was not completely inspected

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The skill is presented as an ElevenLabs speech-to-text tool, but the documentation states that user inputs and local media are actually sent to dLazy-controlled infrastructure for inference and storage. This mismatch is security-relevant because users may disclose audio, prompts, or files under false assumptions about the processor, trust boundary, and data handling path.

Intent-Code Divergence

Medium

• Confidence: 97% confidence
• Finding: The documented examples and error cases reference a required '--prompt' parameter and prompt safety failures that do not match the advertised speech-to-text command options. Such contradictions can cause agents or users to invoke the tool incorrectly, misunderstand what data is being transmitted, or trigger unintended backend behaviors inconsistent with a simple STT workflow.

Whitespace Padding

Medium

• Category: Prompt Injection

Content

## 错误处理

| Code | 错误类型                         | 示例信息                                                                                                                 |
| ---- | -------------------------------- | ------------------------------------------------------------------------------------------------------------------------ |
| 401  | 未授权 (API Key缺失或无效)       | `ok: false, code: "unauthorized", message: "API key is missing or invalid"`                                                              |
| 501  | 缺少必填参数                     | `error: required option '--prompt <prompt>' not specified`                                                               |

• Confidence: 70% confidence
• Finding: Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Natural-Language Policy Violations

Medium

• Confidence: 91% confidence
• Finding: The usage section specifies `--language_code [language_code]` with a default of `zh`, which creates a language preference unless the user explicitly overrides it. Under the policy, forcing a specific language without user opt-in is a natural-language locale violation when no justification or explicit choice flow is provided.

Whitespace Padding

Medium

• Category: Prompt Injection

Content

## Error Handling

| Code | Error Type                         | Example Message                                                                                                          |
| ---- | ---------------------------------- | ------------------------------------------------------------------------------------------------------------------------ |
| 401  | Unauthorized (No API Key)          | `ok: false, code: "unauthorized", message: "API key is missing or invalid"`                                                              |
| 501  | Missing required parameter         | `error: required option '--prompt <prompt>' not specified`                                                              |

• Confidence: 70% confidence
• Finding: Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Natural-Language Policy Violations

Low

• Confidence: 88% confidence
• Finding: The command help specifies `--language_code` with a default of `zh`, which imposes a specific language behavior unless the user overrides it. Under the policy, locale or language constraints should be opt-in or clearly justified; this file does not explain why Chinese is the default for all users.

Static analysis

No suspicious patterns detected.