heartmula

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to clone a repository from https://github.com/HeartMuLa/heartlib.git and download model weights from Hugging Face repositories such as HeartMuLa/HeartMuLaGen.
  • [COMMAND_EXECUTION]: Installation and usage steps involve running shell commands to clone repositories and execute Python scripts from the downloaded source code, specifically python ./examples/run_music_generation.py.
  • [DYNAMIC_EXECUTION]: The documentation requires manual patching of Python source files (modeling_heartmula.py and music_generation.py) within the downloaded library to fix compatibility issues with newer dependencies.
  • [INDIRECT_PROMPT_INJECTION]: 1. Ingestion points: The generation script processes lyrics and genre tags from external text files. 2. Boundary markers: No delimiters or explicit instructions are provided to the model to ignore potential injection attempts in the input text. 3. Capability inventory: The skill is capable of writing generated audio files to the local filesystem. 4. Sanitization: No input validation or sanitization logic is documented in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 07:01 PM
Security Audit — agent-trust-hub — heartmula