ffmpeg剪辑大师

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:35 PM UTC
• Version:
  1.0.0

Overview

This is a coherent local FFmpeg media-editing skill with no evidence of hidden network, credential, or persistence behavior, though users should watch output paths because files may be overwritten.

Install only if you are comfortable letting the skill run local FFmpeg commands on media files you select. Use a dedicated OUTPUT_ROOT or unique output paths, because existing output files can be overwritten, and be aware that local logs may record command details and file paths.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 90% confidence
• Finding: The skill advertises execution of local Python/FFmpeg scripts and use of environment variables, but it does not declare corresponding permissions. This creates a capability/permission mismatch that can bypass user and platform expectations about shell execution and environment access, increasing the risk of unintended file processing or command execution in a local context.

Vague Triggers

Medium

• Confidence: 81% confidence
• Finding: The trigger conditions are overly broad, including phrases like similar-meaning terms and any request involving local video or audio processing. Over-broad activation can cause the skill to run in contexts the user did not clearly intend, exposing local files to processing tools and increasing the chance of unsafe or privacy-impacting actions.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The activation prompt is broad enough to match many generic requests about editing or processing local media, which can cause the skill to be invoked when the user did not specifically want FFmpeg-based local file operations. Over-broad routing is dangerous because it can lead to unintended tool selection, unnecessary file-handling workflows, or user confusion about what actions will be performed locally.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The script invokes ffmpeg with the `-y` flag, which forces overwriting existing output files without confirmation. In a file-manipulation skill that operates on local media, this can cause unintended data loss if the computed or user-supplied output path collides with an existing file, especially in automated agent workflows where users may not see a prompt before destructive actions occur.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The script invokes ffmpeg with the '-y' flag, which forces overwrite of the destination file without prompting. In a local media-editing skill that operates on user-specified paths, this can silently destroy existing files if the output path collides with an important file or is chosen incorrectly by the caller.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The script invokes FFmpeg with the `-y` flag, which forces overwriting the output file without prompting the user. In a multimedia editing skill that operates on local files, this can cause unintended data loss if the computed or user-supplied output path already exists, especially during repeated or automated runs.

VirusTotal

61/61 vendors flagged this skill as clean.