evolink-music
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill communicates with the vendor's API at api.evolink.ai to submit music generation tasks and manage file uploads. These network requests are performed via curl in the evolink-music-gen.sh script and are necessary for the skill's primary functionality.
- [DATA_EXFILTRATION]: The skill's documentation (README.md and SKILL.md) defines an upload_file tool that allows the agent to read local files via a file_path parameter and transmit them to the vendor's API. This capability creates a significant surface for the exfiltration of sensitive local data (such as configuration files or credentials) if an agent is tricked into targeting sensitive paths.
- [PROMPT_INJECTION]: The skill processes untrusted user-provided strings for prompts, lyrics, and style tags, which serves as a surface for indirect prompt injection attacks. This is a particular concern given the skill's ability to read local files and send data to an external API.
- Ingestion points: User-provided prompt, lyrics, title, and style parameters in the generate_music tool; file paths and URLs in the upload_file tool.
- Boundary markers: None identified in the provided skill files.
- Capability inventory: The skill has documented tools for reading local files and performing network requests to a remote API.
- Sanitization: The evolink-music-gen.sh script correctly uses jq to escape user-provided strings when constructing JSON payloads for API requests.
Audit Metadata