The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using the gh CLI for repository management and node for scorecard generation. The dispatch function passes user-provided input as arguments to gh workflow run, which requires the agent to handle variable interpolation securely.
[PROMPT_INJECTION]: The skill ingests data from managed repositories, such as cron-state.json and workflow history, creating a potential surface for indirect prompt injection. This data is incorporated into notifications without explicit sanitization or boundary markers, though this is typical for fleet monitoring tools.
[SAFE]: The skill follows security best practices by using environment variables for authentication and interacting only with the official GitHub API, a trusted service.
[SAFE]: Standard Base64 decoding is used for legitimate retrieval of file contents from the GitHub API.

fleet-control