The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/helm_create.py contains a generate_script function that assembles shell scripts by interpolating a template string (sourced from user-controlled arguments) directly into a file.
[DYNAMIC_EXECUTION]: The generate_script function in scripts/helm_create.py automatically applies executable permissions (chmod 0o755) to the generated scripts, facilitating immediate execution of potentially unvetted code.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied application requirements to generate deployment manifests and helper scripts, creating a surface where malicious instructions in a prompt could influence the generated output.
Ingestion points: User-provided application descriptions, container image details, and configuration requirements processed in SKILL.md and passed to helper scripts.
Boundary markers: No explicit delimiters or safety instructions are used to isolate user-provided data during the generation process.
Capability inventory: The skill possesses the ability to write files, modify file permissions, and execute specific CLI tools (helm and kubectl) as defined in the allowed-tools metadata.
Sanitization: The scripts/helm_create.py tool does not implement validation or escaping for the content written into the generated executable scripts.

generating-helm-charts