flowio

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation and usage patterns for the 'flowio' Python library. All identified patterns are consistent with standard data processing workflows for bioinformatics.
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to install the 'flowio' package using the 'uv' package manager. This is a legitimate and well-known library for handling FCS files.
  • [COMMAND_EXECUTION]: The skill examples demonstrate reading from and writing to local FCS and CSV files. These operations are limited to the intended functionality of the library and do not involve arbitrary command execution or shell injection risks.
  • [DATA_EXFILTRATION]: No network operations or suspicious data transfer patterns were identified. All data processing is performed locally on the user's machine.
  • [PROMPT_INJECTION]: The instructions are purely functional and do not attempt to override agent safety guidelines or manipulate underlying model behavior.
  • [INDIRECT_PROMPT_INJECTION]: While the skill involves parsing external FCS files, the risk is negligible as the skill does not expose high-risk capabilities like network access or system command execution that could be exploited via malicious file content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 06:23 AM