The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill acts as a documentation repository for a well-known open-source biological data framework.
[PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection as it facilitates the ingestion of external data from various sources such as CSV, AnnData, Parquet, and web APIs.
Ingestion points: Data ingestion occurs through standard functions like pd.read_csv, ad.read_h5ad, and requests.get as documented in references/data-management.md and references/integrations.md.
Boundary markers: There are no explicit instructions or delimiters used to separate user data from agent instructions in the provided code examples.
Capability inventory: The framework allows for file system operations via artifact saving and network connectivity via the requests library.
Sanitization: No specific input validation or sanitization of data content for potential prompt injection was found in the documentation snippets.
[EXTERNAL_DOWNLOADS]: The skill provides standard instructions for installing the lamindb package and its ecosystem through the Python package manager pip or uv.
[COMMAND_EXECUTION]: Documentation in references/setup-deployment.md contains standard Linux administration commands using sudo for configuring system-wide cache directories, which is consistent with the skill's purpose as a deployment guide.

lamindb