The Agent Skills Directory

[COMMAND_EXECUTION]: The reference documentation for the BRENDA database (references/brenda.md) explicitly instructs the agent to generate and execute a local Python script using the zeep library. This is required to interact with the database's SOAP API, representing a dynamic code execution pattern.
[EXTERNAL_DOWNLOADS]: The skill is designed to perform API requests to 78 external domains to retrieve data. These include well-known government and academic services such as NASA (api.nasa.gov), the National Institutes of Health (pubchem.ncbi.nlm.nih.gov), the European Bioinformatics Institute (ebi.ac.uk), the US Treasury (fiscaldata.treasury.gov), and the World Bank (api.worldbank.org).
[PROMPT_INJECTION]: By ingesting data from 78 untrusted external databases, the skill has a significant attack surface for indirect prompt injection. Instructions embedded within external database records could potentially influence agent behavior. The skill currently lacks specific guidance for sanitizing or using clear boundary markers for this external data when returning results.

database-lookup