experiment-tracking-swanlab
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides documentation and code examples for SwanLab, an open-source experiment tracking tool, designed for monitoring and visualizing machine learning experiments.
- [EXTERNAL_DOWNLOADS]: The skill recommends installing standard and reputable machine learning packages including swanlab, pillow, soundfile, transformers, pytorch-lightning, and fastai through official package managers.
- [CREDENTIALS_UNSAFE]: The skill correctly demonstrates secure secret management by suggesting the use of environment variables (e.g., using os.environ['SWANLAB_API_KEY']) for API authentication rather than hardcoding sensitive tokens.
- [INDIRECT_PROMPT_INJECTION]: The skill contains data ingestion points where metrics and text are logged via the swanlab.log API (identified in SKILL.md and references/visualization.md). Although explicit boundary markers and sanitization are not detailed, the skill's capabilities are limited to logging and visualization, which does not present a high-risk surface for compromising agent behavior.
Audit Metadata