alterlab-scvelo

Pass

Audited by Gen Agent Trust Hub on Sep 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the scvelo, scanpy, and numpy libraries, which are standard, well-known packages for single-cell data analysis. Installation is recommended via official package registries using the uv tool.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process single-cell RNA-seq data from external file formats such as .loom and .h5ad.
  • Ingestion points: Data loading is performed via scv.read() and sc.read_h5ad() as described in SKILL.md and implemented in scripts/rna_velocity_workflow.py.
  • Boundary markers: None; the data processed consists of structured binary matrices rather than natural language, significantly limiting the risk of instruction injection.
  • Capability inventory: The skill performs local file system operations, including creating output directories (os.makedirs), writing processed data (adata.write_h5ad), and saving visualization plots. It does not perform any network operations.
  • Sanitization: Input validation is handled by the AnnData and scVelo frameworks, which enforce strict schema and type checks on the ingested data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 25, 2026, 02:36 PM