The Agent Skills Directory

CREDENTIALS_UNSAFE (SAFE): The file gitops.md contains a dummy API key (sk-live-123abc...) specifically within a 'Bad Example' section to demonstrate why hardcoding secrets is dangerous. This is educational content and not a functional credential leak.
EXTERNAL_DOWNLOADS (SAFE): The skill references the official @zuplo/runtime package and the zuplo/zuplo-deploy-action GitHub Action. These are standard dependencies for the platform and are used appropriately in configuration examples.
COMMAND_EXECUTION (SAFE): Example code snippets include standard curl commands and CI/CD yaml definitions (GitHub Actions). These are provided as static documentation for deployment workflows and do not trigger unauthorized execution by the agent.
DATA_EXFILTRATION (SAFE): No patterns of data exfiltration were found. The skill emphasizes secure handling of environment variables and the use of encrypted secrets.
PROMPT_INJECTION (SAFE): No instructions targeting agent behavior override or safety bypass were detected. The markdown content follows a standard instructional format.

zuplo