lamindb-data-management
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes biological data artifacts (such as .h5ad and .parquet files), creating a potential surface for indirect prompt injection if those artifacts contain malicious metadata.
- Ingestion points: Artifact.from_anndata, Artifact.from_df, and ad.read_h5ad in SKILL.md.
- Boundary markers: No specific delimiters or safety instructions are used to separate untrusted metadata from instructions.
- Capability inventory: The skill has capabilities to write to the filesystem (artifact.save), access the network (lamin login), and track environment lineage (ln.track).
- Sanitization: Processing relies on the standard parsing behavior of the bioinformatics libraries used.
- [COMMAND_EXECUTION]: The skill provides instructions for standard shell commands used to initialize and configure the LaminDB environment, including login and storage initialization.
- [EXTERNAL_DOWNLOADS]: The skill uses the pip package manager to install necessary libraries including lamindb, pandas, and bionty, which are required for the framework to function.
Audit Metadata