The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Python subprocess module to execute aws s3 CLI commands for listing directories and copying files from the GenomeArk public bucket.
[EXTERNAL_DOWNLOADS]: Fetches genomic assemblies, QC reports, and k-mer histograms from the well-known Amazon Web Services (AWS) S3 service via the public s3://genomeark/ bucket.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, including S3 filenames and text files such as BUSCO and GenomeScope summaries. It parses this untrusted external data to extract metrics using regular expressions and type conversion, which represents a potential injection surface.
Ingestion points: S3 filenames and the content of evaluation report files (e.g., _genomescope__Summary.txt, output_merqury.tabular) in qc-data-fetching.md and assembly-date-extraction.md.
Boundary markers: None identified.
Capability inventory: subprocess.run for executing AWS CLI commands and writing files to the local directory.
Sanitization: Uses regular expression matching, string splitting, and numerical type conversion (int, float) to validate scientific data extracted from external reports.

genomeark-aws