Add interactive engine selection and secret configuration to init command

[Copilot]

Plan: Update init command with interactive mode

• Add interactive mode when init is invoked without arguments

  • Prompt user to select engine (Copilot, Claude, or Codex)
  • If Copilot selected: generate Copilot YAML config and mcp.json
  • For each engine: inspect environment for secrets
  • Recommend and attempt to create secrets using gh command
  • Ask confirmation before configuring secrets
  • Gracefully handle permission errors when setting secrets

• Implementation approach:

  • Modified init_command.go to detect when no engine/flags provided
  • Added InitRepositoryInteractive function using huh library for prompts
  • Reused existing MCP setup functions when Copilot selected
  • Check environment variables for engine-specific secrets
  • Used existing tokens_bootstrap.go logic for secret recommendations
  • Attempted to create secrets with gh secret set command
  • Added confirmation prompt using huh.Confirm before setting secrets
  • Enhanced error handling to detect and report permission issues
  • NEW: Use EngineName constants from constants.go for type safety
  • NEW: Renamed mcp variable to copilotMcp for clarity

• Testing:

  • Add unit tests for interactive mode detection
  • Add tests for precondition checks (CI/test mode blocking)
  • All existing init tests pass
  • Verified non-interactive mode still works with flags
  • Verified help text displays correctly

• Documentation:

  • Updated command help text with interactive mode section
  • Added examples showing interactive usage

• Code Quality:

  • All linting passing
  • Code formatted successfully
  • Build succeeds
  • Using typed constants for better code maintainability

Summary

Successfully implemented interactive mode for the init command with all requested improvements:

1. Interactive engine selection with confirmation prompts
2. Graceful permission error handling
3. Type-safe engine constants usage
4. Clear variable naming (copilotMcp)

Original prompt

Update the init command behavior as follows:

When the command is invoked without arguments, it should enter an interactive mode and prompt the user to select which agent engine to use: Copilot, Claude, or Codex.

If Copilot is selected:

• Generate the Copilot YAML configuration file.
• Create the mcp.json file.

For each engine, inspect the environment for the corresponding secrets. If the required secrets are present:

• Recommend that the user configure them as repository Actions secrets.
• Attempt to create those secrets using the gh command.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[pelikhan]

@copilot ask confirmation before configuring secrets, gracefully handle the case where gh does not have enough permissions to set the secret.

[Copilot]

@copilot ask confirmation before configuring secrets, gracefully handle the case where gh does not have enough permissions to set the secret.

Added in commit f254409:

1. Confirmation prompt: Users are now asked to confirm before configuring secrets using an interactive huh.Confirm prompt
2. Permission error handling: The code now detects 403/Forbidden/permission errors and displays:
   • Specific message: "Insufficient permissions to set {SECRET_NAME}"
   • Helpful guidance: "You may need to grant additional permissions to your GitHub token"
   • Clear distinction between permission errors and other failures

The secret configuration continues with remaining secrets even if one fails due to permissions.

[pelikhan]

@copilot use types and constants defined in constants.go

[Copilot]

Updated in commit fd1ffd0 to use constants.CopilotEngine, constants.ClaudeEngine, and constants.CodexEngine instead of string literals.

[pelikhan]

@copilot rename to copilotMcp

[Copilot]

Renamed to copilotMcp in commit fd1ffd0.