gget
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill interacts with established and reputable scientific databases including Ensembl, UniProt, and NCBI. All network operations are consistent with standard bioinformatics research workflows.
- [SAFE]: Security guidance for the COSMIC and OpenAI modules is robust, explicitly warning against the exposure of credentials in shell history or command-line arguments and promoting the use of environment variables.
- [COMMAND_EXECUTION]: The skill executes specialized bioinformatics binaries (such as Muscle and DIAMOND) via the gget library to perform sequence alignments and local similarity searches, which is its core functionality.
- [EXTERNAL_DOWNLOADS]: Reference genomes, viral sequences, and protein model parameters are retrieved from well-known academic and government repositories (e.g., NCBI, EBI). These downloads are handled through documented library functions.
- [INDIRECT_PROMPT_INJECTION]: The skill processes biological data, creating a potential surface for indirect injection. 1. Ingestion points: Input FASTA files and gene lists processed in scripts/batch_sequence_analysis.py and scripts/enrichment_pipeline.py. 2. Boundary markers: Absent. 3. Capability inventory: Bash execution, file writing, and network queries via gget modules. 4. Sanitization: Absent; the skill relies on library-level parsing of biological formats. This represents a standard analytical surface for genomic tools and is considered safe in this context.
Audit Metadata