alterlab-flowio
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements standard, well-documented functionality for scientific file handling. All external dependencies (flowio, numpy, pandas, flowkit) are legitimate and widely used tools in the bioinformatics and data science communities.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it ingests external binary FCS files containing user-controllable metadata. 1. Ingestion points: FlowData constructor and read_multiple_data_sets function in SKILL.md, references/api_reference.md, and references/workflows.md. 2. Boundary markers: Absent. 3. Capability inventory: Bash execution (Bash(python:), Bash(uv:)) and file-write (create_fcs, write_fcs) in SKILL.md and references/api_reference.md. 4. Sanitization: Absent, relies on the flowio library's internal parsing logic.
Audit Metadata