suno-create

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell commands like curl to download audio files, cp to move files within the workspace, and executes a local Python script suno_browser.py to perform browser automation tasks.
  • [EXTERNAL_DOWNLOADS]: The skill downloads generated songs in MP3 format from Suno's CDN (cdn1.suno.ai). This is an essential function for the music generation service.
  • [DATA_EXFILTRATION]: The skill is designed to send generated audio files and links to a Telegram chat using a message tool. While this is the intended purpose, it involves sending data to an external service.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it searches for and extracts lyrics from external third-party websites and injects this untrusted content directly into the Suno.com interface via browser automation.
  • Ingestion points: External lyrics websites as described in references/lyrics_search.md and searched via DuckDuckGo.
  • Boundary markers: Absent; extracted lyrics are interpolated into automation scripts without specific delimiters or instructions to ignore embedded commands.
  • Capability inventory: Access to shell command execution (curl, cp, python), browser JavaScript evaluation, file system writes, and network communication through a messaging tool.
  • Sanitization: Absent; the skill performs only basic whitespace stripping on the ingested content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 05:19 PM
Security Audit — agent-trust-hub — suno-create