bio-fasta

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [Indirect Prompt Injection] (SAFE): The skill defines patterns for ingesting untrusted sequence data from external files. Ingestion points: File parsing methods such as Bio.SeqIO.parse() and pysam.FastaFile.fetch(). Boundary markers: The skill treats sequence data as structured strings or objects, isolating it from the control flow. Capability inventory: Limited to biological data processing and statistics; no dangerous system-level operations. Sanitization: Includes explicit validation functions like validate_fasta() and validate_fastq() to check file integrity.
  • [Unverifiable Dependencies] (SAFE): The skill utilizes industry-standard libraries (biopython and pysam) from trusted registries.
  • [Data Exposure & Exfiltration] (SAFE): No unauthorized file access or network exfiltration patterns were detected; operations are restricted to the intended bioinformatics context.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:44 PM