Save your best work | Master AI Automation in 4 hours Master AI Automation in 4 hours Course About Ayush Modules Sample chapter Toolbox The Microcap Minute Classroom / Module 02: Talking to Machines: Prompting / Chapter 5 Save your best work Watch first, then read. Same lesson, your pace. What you will learn – Custom instructions: settings that apply to every chat – Projects/Spaces: persistent homes for ongoing work – Your prompt library as a compounding asset Stop re-explaining yourself If you’ve typed “I’m a student, explain simply” more than twice, you’re paying the same tax repeatedly. The fix exists at three levels: Level 1, Custom instructions. Every major AI (ChatGPT, Claude, Kimi, Qwen) has a settings page where you describe yourself once: “I’m a Class 9 student in India. Prefer plain English with short sentences. Use ₹ not $. Ask clarifying questions before big tasks.” Now every new chat starts pre-briefed. Ten minutes of setup, permanent discount on every future prompt. Level 2, Projects / Spaces / Gems. For ongoing work (exam prep, a coding project, college applications), these let you pin instructions + files to a workspace. Everything inside inherits the context: upload your syllabus once, and every conversation in that project already knows it. This is Module 1’s context-window lesson turned into furniture. Level 3, The library that compounds. The deepest level costs nothing: whenever any prompt produces a great result, save it to prompts/ (your cockpit folder). Not screenshots, the actual text, with a one-line note on when to use it. Anatomy of a saved prompt Keep each file in this shape: # exam-planner When to use: before any exam week. — You are a study strategist. Context: my exam dates and subjects are below, plus my daily free hours. Task: build a day-by-day revision plan using active recall (no passive rereading), hardest subjects on high-energy days. Format: markdown table Day | Focus | Method | Done-checkbox. Six months in, you’ll own twenty of these, a personal toolkit tuned to you , portable across every AI on earth, immune to subscription prices. One warning: don’t hoard mediocre prompts. The library’s value is density, not size. If a prompt needed heavy fixing, save the fixed version only. Try it yourself Set custom instructions on all four AIs from your cockpit (same text everywhere). Create one Project (any AI that supports it), upload something real (syllabus, notes), give it standing instructions, test that it remembers inside new chats. Write prompts/exam-planner.txt using the anatomy above, then actually run it this week. You’ve just automated part of your own discipline. Key takeaways – Custom instructions = free permanent prompting; set them once today. – Projects pin files + rules to ongoing work; use one per real project. – A curated prompt library is a portable, compounding asset, save wins immediately. Download the exercise sheet (PDF) Module workbook (PDF) ← Prev: The conversation is the product Module 02 index → Classroom / Module 02: Talking to Machines: Prompting / Chapter 5 Save your best work What you will learn – Custom instructions: settings that apply to every chat – Projects/Spaces: persistent homes for ongoing work – Your prompt library as a compounding asset Stop re-explaining yourself If you’ve typed “I’m a student, explain simply” more than twice, you’re paying the same tax repeatedly. The fix exists at three levels: Level 1, Custom instructions. Every major AI (ChatGPT, Claude, Kimi, Qwen) has a settings page where you describe yourself once: “I’m a Class 9 student in India. Prefer plain English with short sentences. Use ₹ not $. Ask clarifying questions before big tasks.” Now every new chat starts pre-briefed. Ten minutes of setup, permanent discount on every future prompt. Level 2, Projects / Spaces / Gems. For ongoing work (exam prep, a coding project, college applications), these let you pin instructions + files to a workspace. Everything inside inherits the context:
Save your best work
Written by
in