The Agent Skills Directory

[PROMPT_INJECTION]: The skill allows the agent to read and process untrusted genomic data (VCF, BED, BAM) which could contain malicious instructions designed to manipulate the agent. * Ingestion points: pb.read_vcf, pb.scan_bed, and pb.read_bam in SKILL.md and references/file_io.md. * Boundary markers: Absent. * Capability inventory: Bash, Write, and Edit tools. * Sanitization: None described.
[EXTERNAL_DOWNLOADS]: The skill downloads the polars-bio library from PyPI. Evidence: uv pip install "polars-bio==0.31.0" in SKILL.md.
[COMMAND_EXECUTION]: The skill uses the Bash tool for setup and operations. Evidence: allowed-tools includes Bash in SKILL.md frontmatter.

polars-bio