Claude Prompt Library

30 Claude Code Prompts for Plan Mode

30 copy-paste prompts

Name the goal and the files. Claude Code writes a plan, waits for GO, and does not edit first. Not a sneak rewrite and not a week-long roadmap.

In short: This page contains 30 copy-paste ready prompts, organized into 6 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

Louis Corneloup
By Louis Corneloup · Founder, Techpresso
Last updated ·Hand-curated & tested by the AI Academy team

Plan

5 prompts

Plan Only

1/30

✨ What it does

Claude Code writes a plan and waits in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a planner working in Claude Code. <context> I need you to write a plan and wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Files: [FILES] </inputs> <task> 1. Plan for [GOAL] 2. Stay about [FILES] 3. Wait for GO </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not edit.

Steps

2/30

✨ What it does

Claude Code writes numbered steps in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a steps planner working in Claude Code. <context> I need you to numbered steps I can approve. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Files: [FILES] </inputs> <task> 1. Numbered steps 2. Files per step 3. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not start step 1.

Smallest Slice

3/30

✨ What it does

Claude Code plans the smallest slice in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a slice planner working in Claude Code. <context> I need you to smallest slice that works. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Files: [FILES] </inputs> <task> 1. Smallest slice 2. What waits 3. Stay about [FILES] </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: A slice beats a rewrite.

Two Options

4/30

✨ What it does

Claude Code plans two options in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are an options planner working in Claude Code. <context> I need you to two options then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Files: [FILES] </inputs> <task> 1. Two options 2. Tradeoffs from facts I gave 3. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent option C.

Ask First

5/30

✨ What it does

Claude Code asks then waits in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are an ask planner working in Claude Code. <context> I need you to questions then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Unknown: [UNKNOWN] </inputs> <task> 1. Questions from [UNKNOWN] 2. No edits </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: A question beats a wrong plan.

Scope

5 prompts

In Out

6/30

✨ What it does

Claude Code writes in and out of scope in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a scope planner working in Claude Code. <context> I need you to in scope and out of scope. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Out: [OUT] - Files: [FILES] </inputs> <task> 1. In 2. [OUT] 3. Stay about [FILES] </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Out of scope stays out.

File List

7/30

✨ What it does

Claude Code lists files it will touch in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a files planner working in Claude Code. <context> I need you to files I will touch. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Allowed: [FILES] </inputs> <task> 1. List from [ALLOWED] 2. Ask if you need more 3. No edits </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent a helper file.

Off Limits

8/30

✨ What it does

Claude Code plans around off limits in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a fence planner working in Claude Code. <context> I need you to plan honoring off limits. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Off: [OFF] - Files: [FILES] </inputs> <task> 1. Honor [OFF] 2. Plan only </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Off limits stays off.

Deps

9/30

✨ What it does

Claude Code plans packages it needs in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a deps planner working in Claude Code. <context> I need you to packages this needs. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Allowed deps: [DEPS OR NONE] </inputs> <task> 1. Deps from [DEPS] or none 2. Ask before new ones </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent zod.

Data

10/30

✨ What it does

Claude Code plans data it needs in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a data planner working in Claude Code. <context> I need you to data and migrations this needs. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - What I know: [KNOW] </inputs> <task> 1. From [KNOW] 2. UNVERIFIED on missing 3. No edits </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent a table.

Risk

5 prompts

Risks

11/30

✨ What it does

Claude Code plans the risks in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a risk planner working in Claude Code. <context> I need you to plan risks then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Files: [FILES] </inputs> <task> 1. Risks from facts 2. Stay about [FILES] 3. No fake Sev-1 </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent downtime.

Rollback

12/30

✨ What it does

Claude Code plans rollback in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a rollback planner working in Claude Code. <context> I need you to plan rollback then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Ship: [SHIP OR VERIFY] </inputs> <task> 1. Rollback from [SHIP] or ask 2. No edits </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent helm.

Test Plan

13/30

✨ What it does

Claude Code plans the tests in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a test planner working in Claude Code. <context> I need you to plan tests then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Test: [TEST OR VERIFY] </inputs> <task> 1. Tests from [TEST] or ask 2. No edits </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent npm test.

Flag

14/30

✨ What it does

Claude Code plans a feature flag in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a flag planner working in Claude Code. <context> I need you to plan a flag I have. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Flag: [FLAG OR VERIFY] </inputs> <task> 1. Flag or ask 2. No edits </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent FLAG_NEW.

Order

15/30

✨ What it does

Claude Code plans migrate-then-code order in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a order planner working in Claude Code. <context> I need you to plan migrate then code order. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - What I know: [KNOW] </inputs> <task> 1. Order from [KNOW] 2. No edits </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent expand-contract if I did not ask.

These prompts give you the what. Tutorials give you the why.

Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. 300+ tutorials and growing.

Try AI Academy Free

Kinds

5 prompts

Bug Plan

16/30

✨ What it does

Claude Code plans a bugfix in [BUG] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a bug planner working in Claude Code. <context> I need you to plan a bugfix then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [BUG] - Files: [FILES] </inputs> <task> 1. Repro plan 2. Fix plan 3. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not patch yet.

Refactor Plan

17/30

✨ What it does

Claude Code plans a refactor in [SMELL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a refactor planner working in Claude Code. <context> I need you to plan a refactor then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [SMELL] - Files: [FILES] </inputs> <task> 1. Small plan 2. Behavior lock 3. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: A rewrite is out of scope unless I asked.

Feature Plan

18/30

✨ What it does

Claude Code plans a feature in [ASK] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a feature planner working in Claude Code. <context> I need you to plan a feature then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [ASK] - Files: [FILES] </inputs> <task> 1. Slice 2. Out of scope 3. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not start the feature.

Incident Plan

19/30

✨ What it does

Claude Code plans incident debug in [SYMPTOM] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are an incident planner working in Claude Code. <context> I need you to plan incident debug then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [SYMPTOM] - Tried: [TRIED] - Files: [FILES] </inputs> <task> 1. Honor [TRIED] 2. Next checks 3. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent Sev-1.

Docs Plan

20/30

✨ What it does

Claude Code plans the docs in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a docs planner working in Claude Code. <context> I need you to plan docs then wait. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Files: [FILES] </inputs> <task> 1. Docs files 2. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent a wiki.

GO

5 prompts

After GO

21/30

✨ What it does

Claude Code executes the approved plan in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are an after-go planner working in Claude Code. <context> I need you to execute the approved plan only. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Files: [FILES] </inputs> <task> 1. Do only [PLAN] 2. Stay in [FILES] 3. Stop if you need a new step </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not enlarge the plan.

Step One

22/30

✨ What it does

Claude Code does step one only in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a step-one planner working in Claude Code. <context> I need you to do step one only. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Files: [FILES] </inputs> <task> 1. Step one 2. Stop 3. Stay in [FILES] </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not do step two.

Report

23/30

✨ What it does

Claude Code reports after the plan run in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a report planner working in Claude Code. <context> I need you to report after the plan run. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Files: [FILES] </inputs> <task> 1. What you did 2. What you ran 3. What I verify </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Do not invent a pass.

Drift

24/30

✨ What it does

Claude Code stops if the plan drifted in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a drift planner working in Claude Code. <context> I need you to stop if the plan drifted. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Files: [FILES] </inputs> <task> 1. If drift, stop and ask 2. Stay in [FILES] </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: A new idea needs a new GO.

No Extra Files

25/30

✨ What it does

Claude Code skips unplanned files in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a files cutter working in Claude Code. <context> I need you to do not add files the plan skipped. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Files: [FILES] </inputs> <task> 1. Only planned files 2. Ask otherwise </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: A new file needs a GO.

Most people use 10% of Claude. Tutorials unlock the rest.

AI Academy: 300+ hands-on tutorials on Claude, ChatGPT, Midjourney, and 50+ AI tools. New tutorials added every week.

Start Your Free Trial

Fixes

5 prompts

Shorter Plan

26/30

✨ What it does

Claude Code shortens the plan in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a cutter working in Claude Code. <context> I need you to shorter plan. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Cap: [CAP] </inputs> <task> 1. Under [CAP] 2. Keep risks </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Cut the manifesto.

Humanize

27/30

✨ What it does

Claude Code humanizes the plan in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a humanizer working in Claude Code. <context> I need you to humanize the plan. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Keep: [KEEP] </inputs> <task> 1. Cut delve 2. Keep [KEEP] </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: A plan should sound like steps.

Plan Plus Files

28/30

✨ What it does

Claude Code writes a plan plus file list in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a combo planner working in Claude Code. <context> I need you to plan plus file list. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Allowed: [FILES] </inputs> <task> 1. Plan 2. Files </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Files stay a list.

Plan Plus Tests

29/30

✨ What it does

Claude Code writes a plan plus tests in [GOAL] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are a combo planner working in Claude Code. <context> I need you to plan plus test plan. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [GOAL] - Test: [TEST OR VERIFY] </inputs> <task> 1. Plan 2. Tests </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Tests stay runnable or asked.

Unverified Plan

30/30

✨ What it does

Claude Code tags invented plan bits in [PLAN] without wandering outside the files you named. Plan if more than two files change, then tell me what you ran.

You are an honesty planner working in Claude Code. <context> I need you to tag invented plan bits. Stay inside the scope I name. Do not refactor neighbors. Do not add packages I did not ask for. </context> <inputs> - Primary: [PLAN] - Facts: [FACTS] </inputs> <task> 1. Tag extras 2. Wait </task> <constraints> - Plan before you edit if more than two files change. - Do not touch files I mark OFF LIMITS. - Run the test command I gave. If I gave none, ask before inventing one. - Do not invent APIs, env vars, or tickets. </constraints> <format> A short plan, then the edits, then what you ran and what I should verify. </format>

💡

Pro tip: Drop invented commands.

Free tool

Prompt Optimizer

Turn a rough idea into a structured, professional AI prompt.

Try it free →

Frequently Asked Questions

They are written for Claude Code in the terminal: plan, edit, test, and stay inside plan mode. You can paste a shorter version into claude.ai if you are not in a repo.
Each prompt sets scope and non-goals. Still use Plan mode on multi-file work, and read the diff before you accept edits.
It helps. A short CLAUDE.md with stack, test command, and files Claude must not touch keeps these prompts from inventing a second style guide.
The prompts are free to copy. Claude Code needs your Anthropic login and uses your plan's usage. Local files never leave your machine except through that session.
The hub is a 21-prompt overview. This page is only plan mode: thirty copy-paste jobs with scope, verify, and non-goals already written.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.