bio-phylogenomics

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill employs a local Python helper script (scripts/run_phylogenomics.py) to manage a pipeline of bioinformatics tools, including MAFFT, trimAl, IQ-TREE 3, and VeryFastTree. The script constructs and runs shell commands using subprocess.run based on input manifests.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted genomic sequence data and phylogenetic tree files that could contain adversarial content designed to influence the agent's behavior.
  • Ingestion points: The agent reads markers.tsv, FASTA sequences, and Newick-formatted tree files (.nwk, .treefile) during analysis and post-processing.
  • Boundary markers: The skill does not explicitly define or use delimiters (like XML tags or markdown blocks) to isolate sequence data or tree labels from the rest of the agent's instructions.
  • Capability inventory: The skill can execute specific command-line tools and write various TSV, JSON, and report files to the filesystem.
  • Sanitization: The script implements checksum validation for reference sequences to ensure data integrity before processing.
  • [EXTERNAL_DOWNLOADS]: The skill instructions involve fetching genome and proteome sequences from external biological databases (e.g., NCBI) to support comparative analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 02:46 PM