The Agent Skills Directory

[SAFE]: The skill utilizes official medicinal chemistry libraries (medchem, datamol, rdkit) available on PyPI. The code snippets and provided utility script (scripts/filter_molecules.py) follow standard data science and chemoinformatics practices.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it processes external molecular data files (CSV, SDF, TXT) through the scripts/filter_molecules.py script.
Ingestion points: scripts/filter_molecules.py via the load_molecules function.
Boundary markers: Absent in the data processing utility.
Capability inventory: The skill uses Bash, Write, and Edit tools to process libraries and write output CSV/TXT files.
Sanitization: Input data is parsed through RDKit and Datamol molecular validators, which effectively filters out non-chemical payloads and mitigates common injection risks.
[COMMAND_EXECUTION]: The skill provides a batch filtering script (scripts/filter_molecules.py) intended for use with the Bash tool. The script uses standard argparse for parameter handling and does not perform unsafe shell interpolation or execute arbitrary user-provided commands.

medchem