# Plan before you code > Use Empryo plan mode for multi-file refactors: the agent researches, writes a light or full plan, waits for approval, then executes step by step. For anything touching more than 2-3 files, use plan mode. The agent researches first, writes a structured plan, waits for your approval, then executes step-by-step. ## Start a plan ``` /plan refactor the auth module to use JWT refresh tokens ``` Or switch to plan mode globally: ``` /mode plan ``` In plan mode the agent has its full tool belt, it can read, search, and run read-only commands to make the plan true, but it is instructed to research only and never to change the project. Changing anything is your call: the agent calls the `plan` tool and the plan appears in the sidebar for review. Approval gates (destructive commands, writes outside the project) still apply as normal. ## Review the plan - **Approve** - switches the lane to auto and executes the plan step-by-step, with progress in the sidebar. - **Revise** - type what you want changed, the agent rewrites. - **Cancel** - discards the plan. ## Plan types - **Light plan** - a checklist, no code. Fast, low context. - **Full plan** - includes code snippets and old→new diffs. More context, higher confidence on big refactors. The agent picks based on the request. You can override in the prompt ("make a detailed plan"). ## Why plan mode For small fixes, planning is overhead. For multi-file refactors, it catches mistakes before you commit to them - cheaper than re-running an expensive edit session. ## Resume a plan If you close Empryo mid-plan, `/session continue` picks up where you left off. Plans are saved to `.empryo/plans/` so they survive restarts.