The Agent Skills Directory

[SAFE]: The skill consists of markdown instructions and references that define a methodology for agent coordination. It does not contain executable code, installers, or remote script downloads.
[SAFE]: The orchestration logic includes built-in security constraints, explicitly stating that leading a team does not grant permission to act on external systems or change scope without authorization.
[SAFE]: The skill provides clear guidance on secure credential handling, instructing agents to follow established project procedures (like 1Password) and specifically warns against weakening production authentication for convenience during testing.
[SAFE]: Data hygiene is emphasized through mandatory cleanup instructions for fixture sessions and secret files created during QA and recording milestones.
[SAFE]: The instructions for agent handoffs and reporting include structured boundary markers, such as specific task brief fields and completion statuses, to help maintain context integrity when processing reports from different agents.

agent-orchestration