can-opener
Warn
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple shell commands such as
mix,cp,grep, andrm. Several of these commands use variables (e.g.,<package_name>,<spec_path>) that are filled by user input. This pattern is susceptible to command injection if the inputs contain shell metacharacters (e.g.,;,&,|).\n- [REMOTE_CODE_EXECUTION]: The skill uses themix run -ecommand to execute Elixir code dynamically, incorporating the user-provided<ModuleName>. It also performs amix deps.getwhich fetches and executes code from the vendor's GitHub repository (github.com/agoodway/can_opener.git).\n- [EXTERNAL_DOWNLOADS]: The skill downloads Elixir dependencies from GitHub. It also provides an option to fetch an OpenAPI specification from a remote URL provided by the user, which involves downloading external content to the local system.
Audit Metadata