OpenSpec/config/profile

• openspec config profile
  • This is a Curses-style interactive config wizard. I'm not sure why it's called "profile"
  • OpenSpec/config/profile/Step 1 - Delivery and Workflows

    • This is a bit like a UX/Pattern/Table of Contents that enables one to select whether subsequent steps of the openspec config profile will configure OpenSpec/Delivery, OpenSpec/Workflows, or both

      • openspec config profile
        
        Current profile settings
          Delivery: both
          Workflows: 4 selected (core)
          Delivery = where workflows are installed (skills, commands, or both)
          Workflows = which actions are available (propose, explore, apply, etc.)
        
        ? What do you want to configure?
        ❯ Delivery and workflows
          Delivery only
          Workflows only
          Keep current settings (exit)
        
        Update install mode and available actions together
        ↑↓ navigate • ⏎ select
        Config profile cancelled.
        

  • OpenSpec/config/profile/Step 2 - Configure Delivery

    • openspec config profile
      
      Current profile settings
        Delivery: both
        Workflows: 11 selected (custom)
        Delivery = where workflows are installed (skills, commands, or both)
        Workflows = which actions are available (propose, explore, apply, etc.)
      
      ✔ What do you want to configure? Delivery only
      ? Delivery mode (how workflows are installed):
      ❯ Both (skills + commands) [current]
        Skills only
        Commands only
      
      Install workflows as both skills and slash commands
      ↑↓ navigate • ⏎ select
      

    • There are three options:
      • 1 - Both
      • 2 - opsx installed in project as Agent/Skills
      • 3 - opsx installed in project as UX/Pattern/Text Input/Slash Command
    • Why might we want both?
      • Agent/Command/vs Agent Skills

        • AI Agent Slash Comands vs AI Agent Skills

          • tldr
            • An AI/Agent can invoke Agent/Skills when it determines them to be relevant according to their description.
            • An AI/Agent will only invoke a Agent/Command if the human operator invokes it, or, in some cases, if the prompt instructions for a skill say to invoke that slash command.
          • In Claude Code, Agent/Skills can invoke Agent Commands. In other platforms, it may or may not be the same.
          • See Anthropic's docs comparing agent slash commands with Agent/Skills in Claude/Code/Command/Slash/Docs/Comparison with Skills
  • OpenSpec/config/profile/Step 3 - Configure Workflow