docker-containerization
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION] (MEDIUM): The scripts
scripts/docker-build.shandscripts/docker-run.shuse theevalfunction to execute dynamically constructed shell commands. - Evidence: In
scripts/docker-build.sh, the variableDOCKER_CMDis built from user-influenced arguments like-n,-t, and-b. The script then executeseval $DOCKER_CMD. - Risk: If the agent is tricked via indirect prompt injection into passing arguments containing shell metacharacters (e.g.,
;,&&, or backticks), it could lead to arbitrary command execution on the host machine. - [DATA_EXPOSURE & EXFILTRATION] (SAFE): No hardcoded credentials, sensitive file access, or unauthorized network operations were detected. The scripts handle environment files and registry URLs as expected for their stated purpose.
- [UNVERIFIABLE DEPENDENCIES] (SAFE): The
package.jsonfile contains no external dependencies, and the skill does not attempt to download or execute remote scripts during runtime.
Audit Metadata