The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates shell commands using variables derived from the local file system, such as sops --decrypt <file>.enc.yaml. This interpolation of workspace-derived file names into a shell context is a potential command injection vector if the names contain malicious shell characters.
[COMMAND_EXECUTION]: The workflow executes Python scripts located in a relative path outside the current skill directory (../sops-setup/scripts/). While these are vendor-owned resources, the execution of external scripts is a significant capability.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes content from .enc.yaml files within the workspace.
Ingestion points: Data enters the agent context through encrypted YAML files found in the project root.
Boundary markers: No specific delimiters or instructions to ignore embedded content are provided in the decryption workflow.
Capability inventory: The skill has access to shell execution (sops), file deletion (rm), and script execution (python3).
Sanitization: The workflow does not explicitly describe sanitization or validation of the decrypted content before further processing or conversion.
[SAFE]: The skill implements security best practices by verifying encryption key presence before operation and instructing users to exclude decrypted secrets from version control using .gitignore.

sops-decrypt