The Agent Skills Directory

[SAFE]: The skill interacts with the official BRENDA Enzyme Database via its SOAP API (brenda-enzymes.org). This is an established and well-known scientific resource, and all network operations are directed to this legitimate service domain.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits a standard surface for processing external data retrieved from a database.
Ingestion points: Data is retrieved from the BRENDA database in SKILL.md via client.service methods such as getKmValue, getTurnoverNumber, and getEcNumber.
Boundary markers: No specific delimiters or "ignore instructions" markers are used in the provided Python examples when processing database responses.
Capability inventory: The skill performs network operations for database queries and includes examples of local file writing (e.g., df.to_csv in Workflow 1 of SKILL.md).
Sanitization: There is no evidence of sanitization or content filtering for the retrieved scientific data before it is processed or stored.
Assessment: This surface is a functional requirement for scientific data integration and is assessed as safe given the nature of the data and the reputation of the source.
[SAFE]: The skill documentation demonstrates security awareness by recommending that users store API credentials in environment variables rather than hardcoding them into scripts.

brenda-database