The Agent Skills Directory

[SAFE]: The skill provides documentation and code examples for the pyopenms library, which is a well-established tool in the bioinformatics community for mass spectrometry analysis. The code snippets demonstrate standard scientific data handling practices.
[EXTERNAL_DOWNLOADS]: The skill recommends installing the pyopenms package through a standard package manager (uv pip install). The referenced resources, including documentation at pyopenms.readthedocs.io and the official repository at github.com/OpenMS/OpenMS, are the legitimate project sites.
[INDIRECT_PROMPT_INJECTION]: The skill involves processing scientific data files (e.g., mzML, mzXML, FASTA). While this is an ingestion surface for external data, the formats are structured for mass spectrometry peaks and protein sequences rather than natural language instructions, and the library uses specialized scientific parsers that do not present a risk of prompt injection.
Ingestion points: Files loaded via ms.MzMLFile().load(), ms.FASTAFile().load(), and other format-specific loaders in references/file_io.md.
Boundary markers: None (typical for binary/XML scientific data formats).
Capability inventory: File system writing is supported via ms.MzMLFile().store() and related methods.
Sanitization: Not applicable for scientific data formats of this type.

pyopenms