The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a deployment script (scripts/deploy.sh) designed to automate application releases using industry-standard tools including Docker (docker build and docker push), Kubernetes (kubectl), AWS CLI (aws ecs), and Google Cloud SDK (gcloud run).
[DYNAMIC_EXECUTION]: Multiple Python scripts (e.g., scripts/scaffold_api.py, scripts/generate_model.py, scripts/setup_auth.py) generate project structures, source code, and configuration files based on internal templates. This is a functional requirement for scaffolding and does not involve the execution of external, untrusted code.
[PRIVILEGE_ESCALATION]: Documentation in references/troubleshooting.md mentions using sudo to resolve file permission errors during script execution, while explicitly acknowledging it as a non-recommended approach.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it generates code based on user-provided inputs.
Ingestion points: Configuration parameters and schema files provided to the Python scaffolding scripts in scripts/.
Boundary markers: Absent within the generated code templates.
Capability inventory: File system write operations across all scaffolding scripts and shell command execution in scripts/deploy.sh.
Sanitization: Input validation is limited to basic type checking and choice selection via the argparse and json libraries.

backend-developer