The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references a third-party debug container image (ghcr.io/bikramkgupta/do-app-debug-container-python) for infrastructure diagnostics. While documented, this image is hosted under an individual's GitHub account rather than an official organization repository.
[COMMAND_EXECUTION]: The skill provides numerous templates and instructions to execute administrative commands using doctl and the GitHub CLI (gh). These commands manage cloud resources, repository environments, and CI/CD secrets.
[DATA_EXFILTRATION]: The skill manages sensitive configuration including DIGITALOCEAN_ACCESS_TOKEN and DATABASE_URL. It follows security best practices by using the gh CLI to set environment secrets, which ensures the AI agent only generates the command structure without ever accessing the actual credential values.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection attack surface because it processes untrusted application data during the deployment workflow.
Ingestion points: Reads application specifications from .do/app.yaml and processes deployment logs via doctl apps logs.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the workflow templates.
Capability inventory: The skill has the authority to create, update, and delete cloud applications and modify repository secrets.
Sanitization: There is no evidence of sanitization or schema validation performed on the app specification or log data before it is interpreted by the agent.

deployment