tooluniverse-metabolomics-analysis
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected. The skill instructions and code examples adhere to standard bioinformatics practices.- [COMMAND_EXECUTION]: The skill provides Python code for data loading, quality control, normalization, and statistical analysis using well-known scientific libraries (pandas, numpy, scipy, sklearn, and statsmodels).- [DATA_EXFILTRATION]: No exfiltration patterns or sensitive file access were identified. All operations are local or utilize the platform's internal tool SDK for database matching.- [PROMPT_INJECTION]: The skill was evaluated for indirect prompt injection surface. 1. Ingestion points: Data is ingested from user-provided CSV/mzML files via load_metabolomics_data in code_examples.md. 2. Boundary markers: No explicit delimiters or instructions to ignore embedded content are present. 3. Capability inventory: The skill uses tooluniverse.run_one_function for metabolite identification and enrichment. 4. Sanitization: No sanitization is implemented for the content of processed data files. This represents a standard surface for biological data analysis tools with no elevated risk.
Audit Metadata