The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill includes functionality to download demonstration datasets.
Evidence: The script scripts/rna_velocity_workflow.py uses scv.datasets.pancreas() to fetch sample data from the official scVelo repository for testing and demonstration purposes.
[INDIRECT_PROMPT_INJECTION]: The skill processes external biological data files, which is a standard requirement for bioinformatics tools but introduces a theoretical attack surface for malicious metadata.
Ingestion points: Data is loaded from .loom and .h5ad files using scv.read() and sc.read_h5ad() in SKILL.md and the workflow script.
Boundary markers: No explicit instruction delimiters or 'ignore' instructions are used when interpolating file metadata into the agent's context.
Capability inventory: The skill can perform file system operations such as writing processed AnnData objects (adata.write_h5ad) and saving visualization images. It does not have capabilities for arbitrary command execution or unauthorized network data exfiltration.
Sanitization: The skill relies on the standard parsing capabilities of the scvelo and scanpy libraries.

scvelo