Working well with Enjoy

Habits that help agents do their best work for you.

Set up a new project for smooth workA few minutes of setup on a new project pays off in every conversation after it. Turn on version history, write a short brief, add instructions, and make sure the agent can check its work.
Describe what you wantAgents do their best work when a request says what you want, why, what to keep in mind, and how you'll know it's done. Screenshots and references fill in the rest.
Use one conversation per taskGive each task its own conversation, and keep follow-ups about that task in it. Agents stay focused, and your conversation list reads like a to-do list.
Plan before you buildFor bigger changes, have the agent explore and propose a plan before it touches anything. You catch misunderstandings while they're cheap to fix.
Check the agent's workAn agent stops when the work looks done to it. Ask it to test its changes and show you proof, then try the result yourself before you move on.
Give feedback the agent can act onSay what you saw, what you expected, and where. Correct course early, even while the agent is still working, and answer its questions instead of talking past them.
Mark finished work as completedWhen a task is done, mark its conversation completed instead of adding the next task to it. Your list shows what's finished at a glance, and nothing is lost.
Start fresh when a conversation gets longLong conversations make agents forget early details. Learn the signs, and how to move to a new conversation without losing what you've built up.
What to do when an agent goes in circlesIf an agent keeps trying fixes that don't work, more of the same rarely helps. Stop it, get a diagnosis, and restart with what you learned, a plan, or a different agent.
Keep project knowledge in docsA new conversation doesn't carry over what was said in earlier ones. Write down decisions, plans, and how your project works in docs, so any agent can pick them up when it needs them.
Write instructions your agent will followProject instructions reach the agent in every conversation, so you stop repeating yourself. Here's what to put in them, what to leave out, and how to keep them working.
Turn repeat work into workflowsWhen you catch yourself sending the same kind of request again, save it as a workflow. Run it in one step, add it to a message, and schedule it once it's reliable.
Try ideas without risking your projectAsk your agent to save working versions as commits, and try bigger ideas in a worktree, a separate copy you can merge in or throw away.
Run agents in parallelSeveral conversations can work at the same time, each with its own agent. Give them separate tasks, keep overlapping changes in worktrees, and check in when they need you.
Stay in the loop without watchingYou don't need to watch an agent work. Let Enjoy tell you when it replies or needs you, answer questions quickly, and plan around the steps that need your computer.
Match the effort level to the taskHigher effort makes an agent think longer, which helps on hard problems but is slower and uses more of your plan. Learn when to turn it up and when to turn it down.
Make your agent usage lastYour AI subscription has usage limits. A few habits stretch them further, and Enjoy shows how much you have left and helps you keep going when you hit one.
Keep your projects tidyA few minutes of tidying keeps Enjoy quick to scan. Pin the projects you're working on, close out finished conversations, pin key docs, and archive what's stale.
Work from your phoneYour phone steers agents that run on your computer. It's great for checking in, answering questions, and quick follow-ups. Set things up before you leave, and save a few jobs for the computer.
Work well with a teammateWhen someone joins your computer, their requests run on your agents and your AI plan. Split the work, give the right access, and keep shared knowledge where both of your agents can find it.
Learn as you goYour agent can explain your project, its own changes, and the words it uses. A few questions along the way make you a better director of its work.
Keep passwords and API keys safeAnything you type to an agent goes to your AI provider and is saved with the conversation. Keep secrets in a file instead, tell the agent where they are, and handle sign-ins yourself.