scvelo

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external biological data files (.loom, .h5ad), which constitutes a potential surface for indirect prompt injection.
  • Ingestion points: Data ingestion occurs through scv.read and sc.read_h5ad in SKILL.md and scripts/rna_velocity_workflow.py.
  • Boundary markers: The skill lacks explicit boundary markers or instructions to isolate processed data from the agent's logic.
  • Capability inventory: The skill possesses file-writing capabilities through adata.write_h5ad and data visualization functions.
  • Sanitization: No specific sanitization or validation is applied to the content of the ingested data files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 01:46 AM
Security Audit — agent-trust-hub — scvelo