setup-liana-remote-tool
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads and installs a Python package (
tuplatform-connect) fromhttps://connect.aiscientist.tools. Although this is a project-specific resource provided by the author, it is an external executable dependency. The skill mitigates this by using a SHA-256 hash pin for integrity verification. - [PROMPT_INJECTION]: The skill demonstrates an indirect prompt injection surface.
- Ingestion points: The skill is designed to process external biological data files (e.g.,
tiny.h5ad) at runtime. - Boundary markers: There are no explicit delimiters or warnings instructed to the agent to disregard potential instructions embedded within the processed data files.
- Capability inventory: The skill possesses the capability to launch network servers (
tu serve) and execute Python modules as subprocesses. - Sanitization: While the skill mentions data normalization, it does not include sanitization steps to prevent malicious natural language instructions within the data from influencing the agent.
- [COMMAND_EXECUTION]: The skill requires the execution of multiple shell commands to manage virtual environments, install dependencies, and run validation scripts. This includes the execution of inline Python scripts using heredocs.
- [SAFE]: The skill adheres to secure development practices for credential handling. It explicitly warns against providing sensitive tokens in shell arguments and recommends using restricted environment variables or local files with secure permissions (0600).
Audit Metadata