The Agent Skills Directory

[SAFE]: The skill is designed for genomic research and implements robust, standard procedures for interacting with the Ensembl database. The provided Python script and documentation align with legitimate scientific use cases, incorporating appropriate error handling for API responses.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external data from the Ensembl REST API into the agent's context, which constitutes an indirect prompt injection surface. Given that the source is a well-known international scientific organization (EMBL-EBI) and the data is highly structured biological information, the security risk is negligible.
Ingestion points: Data is retrieved via the Ensembl REST API using the requests library in both the scripts/ensembl_query.py client and the examples in SKILL.md.
Boundary markers: The skill does not implement specific boundary markers or delimiters for the API output; results are returned as JSON or raw text.
Capability inventory: The skill possesses network communication capabilities via the requests library to query rest.ensembl.org and grch37.rest.ensembl.org.
Sanitization: There is no evidence of specific sanitization or filtering logic designed to detect or neutralize natural language instructions embedded within the retrieved genomic data.
[EXTERNAL_DOWNLOADS]: The skill communicates with official Ensembl infrastructure to download genetic sequences, variant predictions, and annotations. These network operations are standard for the tool's primary purpose and target established scientific repositories.

ensembl-database