Advanced Features (AF)

Weight: 10% — measures whether you leverage the full capabilities of the AI coding environment.

Delegation (0–10)

Do you use subagents, skills, or parallel execution?

High delegation:

• Uses subagents for independent tasks (e.g., research + implementation in parallel)
• Leverages skills for specialized operations
• Delegates repetitive tasks to automated processes
• Uses parallel tool calls when tasks are independent

Low delegation:

• Does everything sequentially in the main conversation
• Never uses subagents or skills
• Manually performs repetitive tasks that could be automated

Configuration (0–10)

Do you configure the AI environment for your project?

High configuration:

• Maintains a CLAUDE.md file with project context, conventions, and instructions
• Uses hooks for automated checks (pre-commit, test runners)
• Uses plan mode for complex tasks
• Configures MCP servers for external integrations

Low configuration:

• No CLAUDE.md — the AI has no project context
• No hooks configured
• Never uses plan mode, even for complex tasks
• Doesn’t leverage available integrations

Improving AF

1. Create a CLAUDE.md — include project structure, coding conventions, and common commands
2. Use plan mode — for multi-file changes, plan before implementing
3. Try subagents — delegate independent research or exploration tasks
4. Set up hooks — automate testing, linting, or formatting on file changes
5. Configure MCP — connect external tools (databases, APIs, project management)