The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external biological data files such as CSV and H5AD (AnnData). While these files are standard in scientific workflows, they represent an attack surface where malicious metadata could be embedded to influence an agent's interpretation of the results.
Ingestion points: pd.read_csv in examples/cassiopeia_reconstruction.py and sc.read_h5ad in examples/cospar_dynamics.py.
Boundary markers: No specific delimiters or "ignore instructions" warnings are used in the data loading process.
Capability inventory: The skill performs local data processing, phylogenetic tree reconstruction, and visualization. It does not include network exfiltration or sensitive file system write capabilities in the provided scripts.
Sanitization: The example scripts do not implement validation or sanitization of the input data structure.
[UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill documentation and examples recommend the installation of standard, well-known scientific libraries including cassiopeia-lineage, cospar, scanpy, and numpy. These are reputable packages from the bioinformatics community and their usage is consistent with the skill's stated purpose.

bio-single-cell-lineage-tracing