What does it mean that scope optimization?

An agent that tries to do too much does nothing well. Optimize by narrowing scope to what matters.

Example: You have a morning routine agent — an automatic behavioral pattern that is supposed to prepare you for a productive day. Over time, the scope of this agent has expanded. It now includes: checking email, scanning three news sites, reviewing your calendar, updating your task list, doing a ten-minute workout, meditating for five minutes, journaling, reviewing your goals, preparing a healthy breakfast, tidying the kitchen, watering the plants, and checking your social media metrics. The agent fires every morning, but it takes two and a half hours, you rarely complete all twelve steps, and you start most days feeling behind rather than prepared. You decide to optimize scope. You ask: which of these steps serve the agent's core purpose — preparing me for a productive day? Email, news, and social media metrics generate anxiety and reactivity. Plant watering and kitchen tidying are maintenance tasks, not preparation. You narrow the agent to five steps: calendar review, task list check, ten-minute workout, five-minute meditation, breakfast. The narrowed agent takes forty-five minutes, runs to completion every morning, and actually achieves its purpose. The other tasks do not disappear — they get reassigned to agents whose scope they belong in.

Try this: Select one agent — a habit, routine, workflow, or recurring process — that currently feels bloated or unreliable. List every action this agent currently includes. For each action, classify it as core (directly serves the agent's primary purpose), supporting (indirectly useful but not essential), or migrated (belongs in a different agent or process). Remove all migrated items immediately. For supporting items, ask: does this item's presence make the core items less likely to execute? If yes, remove it. Rewrite the agent's scope as a single sentence: 'This agent does X.' If you cannot state the scope in one sentence, the scope is still too broad. Run the narrowed agent for one week and record completion rates and subjective effectiveness compared to the bloated version.