The Agent Skills Directory

[SAFE]: The skill acts as a technical documentation and style guide for the Nika workflow language, promoting secure development lifecycle practices.
[COMMAND_EXECUTION]: While the skill describes how to use shell commands via the exec: verb, it mandates a 'Native-first' policy that requires authors to exhaust built-in tools (like nika:read, nika:write, and nika:fetch) before resorting to shell execution. It also requires an 'Exec ledger' to justify any remaining shell commands.
[CREDENTIALS_UNSAFE]: The instructions explicitly promote secure secret handling, directing users to use environment or secret variables (${{ secrets.X }}) and forbidding the inlining of credentials or their exposure in command arguments.
[DATA_EXFILTRATION]: The skill provides guidance on using nika:fetch for network operations but requires explicit declaration of egress sinks and secret masking in headers, which are standard security controls for workflow engines.

nika-authoring