aave
Pass
Audited by Gen Agent Trust Hub on Mar 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes
npxto fetch and run thefibxpackage from the NPM registry at runtime. The use of the@latesttag ensures the most recent version is used but introduces a dependency on external code hosted on a public repository. This package is identified as a resource provided by the vendor.\n- [COMMAND_EXECUTION]: The skill is designed to execute system commands viaBashto interact with the Aave protocol through thefibxCLI. These commands perform critical financial actions such as supplying, borrowing, and repaying assets.\n- [INDIRECT_PROMPT_INJECTION]: The skill logic processes the output of CLI commands likefibx aave statusto assess account health factors and determine whether to proceed with transactions, which establishes a data-driven logic flow based on external output.\n - Ingestion points: Account health and balance data returned from the
fibxCLI commands.\n - Boundary markers: No specific delimiters or instructions are provided to the agent to treat tool output as untrusted or to ignore embedded instructions.\n
- Capability inventory: The skill has the authority to execute financial transactions and system-level commands based on its interpretation of the tool data.\n
- Sanitization: There is no evidence of explicit validation or sanitization of the CLI output before it is processed by the agent's logic.
Audit Metadata