elixir
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [Prompt Injection] (SAFE): No patterns of instruction override, jailbreaking, or system prompt extraction were detected. The instructions are focused on coding best practices.
- [Data Exposure & Exfiltration] (SAFE): There are no hardcoded credentials, sensitive file path references, or network exfiltration patterns.
- [Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill does not attempt to install external packages or execute scripts from remote URLs. Mentions of Elixir tools (Ecto, Phoenix) are contextual and not execution commands.
- [Command Execution] (SAFE): No shell command execution or subprocess spawning is present in the skill content.
- [Indirect Prompt Injection] (SAFE): While the skill is designed to interact with code (untrusted data), it lacks dangerous capabilities such as file-writing or network access that would create an exploitable surface.
Audit Metadata