The Agent Skills Directory

[SAFE]: The skill is designed for the legitimate purpose of scaffolding Helm charts and follows best practices for Kubernetes resource management.
[COMMAND_EXECUTION]: The skill uses local shell scripts (scripts/generate_chart_structure.sh and scripts/generate_standard_helpers.sh) to automate file creation. These scripts include robust input validation, such as checking chart names for DNS-1123 compliance, which prevents common injection vulnerabilities.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: User-provided requirements for chart name, container image, and ports in SKILL.md (Stage 1).
Boundary markers: The skill does not explicitly use delimiters for user-provided strings in its internal logic, but the output is structured YAML for Helm.
Capability inventory: File system writes (cat >) and shell execution via bash for scaffolding scripts.
Sanitization: The skill implements strict regex validation in scripts/generate_chart_structure.sh (e.g., validate_chart_name and validate_port) to ensure user inputs are sanitized before being used in commands or template generation. Given its primary purpose as a generator, the residual risk is considered negligible.

helm-generator