Audio Reply

• Outcome: Pass
• Latest audit: May 27, 2026, 9:03 PM
• Version:
  1.1.0

Overview

This skill is a disclosed local text-to-speech helper with reasonable safety guidance; its main caveat is that some trigger phrases are broad and may play audio unexpectedly.

Install this only if you want local audio playback from your agent. Use public, non-sensitive URLs or pasted redacted text, expect uv/model downloads on first use, and be aware that short phrases like "speak" or "say it" may trigger audio generation.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• 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 (4)

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The trigger phrase "talk to me [topic]" is broad conversational language that can easily appear in normal interaction, causing the skill to activate when the user did not explicitly intend to invoke TTS behavior. In an agent environment, overly generic triggers increase the chance of accidental execution and can route arbitrary topics into the skill's fetch/generation pipeline unexpectedly.

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: Single words and short phrases like "speak", "say it", and "voice reply" are highly ambiguous and likely to occur in unrelated conversations, making unintended skill activation more probable. This can cause unauthorized or surprising audio generation, especially if the surrounding system automatically binds these phrases to tool execution.

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: The metadata advertises very broad trigger phrases like "speak" and "say it," which are common in normal conversation and can cause accidental invocation of the skill. In a skill that can fetch URLs and invoke external tooling, unintended activation increases the chance of unwanted network access or audio playback.

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The documented trigger list includes vague standalone phrases without contextual constraints, which makes unintended activation likely. Because the skill can fetch remote content and run TTS commands, false activations can produce unrequested side effects beyond a simple text response.

VirusTotal

63/63 vendors flagged this skill as clean.