5 Claude Prompts for Planning You Can Use Today

These five prompts cover the most common planning conversations. Each is copy-pasteable, with [brackets] where you fill in your specifics.

The prompts are ordered by frequency of use, not complexity.

Prompt 1: The Daily Planning Check-in

Use this: Every morning, 3–5 minutes.

Daily planning check-in. Today is [date].

My plan for today: [paste your tasks or today's row from your weekly plan]
What's changed since I made that plan: [any new fires, reschedules, energy shifts]
My energy right now: [1–10, or a word like "low / medium / sharp"]

What should I actually focus on this morning? Give me three priorities in order. 
Tell me the one thing I'm most likely to avoid that I shouldn't.

Why it works: The “energy right now” input stops Claude from planning your cognitive peaks for low-importance tasks. The final question forces Claude to reason about your behavior, not just your task list — which is where planning conversations get genuinely useful.

Prompt 2: The Weekly Plan

Use this: Sunday evening or Monday morning, 15–20 minutes.

Weekly plan time. Today is [date].

My calendar this week:
[paste your calendar events with rough durations]

Key deadlines this week: [list]
Carry-over from last week I haven't finished: [list]
Energy forecast: [when are you sharpest? Any constraints like travel or illness?]

Build me a weekly plan as an Artifact. One row per day. Columns: Focus Block | Priority Tasks | Admin/Comms | Flags.

In the Flags column: note overloaded days, dependency conflicts, and anything I'll likely push that I shouldn't. Don't tell me to "be realistic" — flag the specific problems you see.

Why it works: The explicit Artifact request produces a formatted table rather than prose. The “Flags” column makes Claude’s concerns visible in the plan itself, not buried in a paragraph below it.

Prompt 3: The Project Decomposition

Use this: When starting a new project, or when a project has stalled.

Project decomposition. I need to complete [project name] by [deadline].

Here's what I know: [2–4 sentences describing the project and its goal]
Known constraints: [time budget, who else is involved, any skill gaps]
What I'm uncertain about: [the parts that feel unclear or risky]

Break this into 4–6 milestones. Format as an Artifact table with columns:
Milestone | Estimated effort | Dependencies | Most important decision before starting

For the "most important decision" column: tell me the thing I need to decide (not do) before each milestone can begin.

Why it works: The “most important decision” column is the key. It forces the planning conversation to surface the deciding that needs to happen before the doing — which is where project delays actually originate.

Prompt 4: The Priority Triage

Use this: When you have too many competing demands.

Priority triage. I have too much to do.

Here's everything competing for my attention: 
[paste full task list — don't filter, dump everything]

My constraints: [hours available today, energy level, any immovable commitments]
Who is waiting on what: [brief — which tasks have external stakeholders expecting something]

Score each item on: urgency × importance × only-I-can-do-this.

Then:
1. Give me the ranked order
2. Explain the two hardest ranking calls you made and why
3. Tell me if anything on this list should be delegated or dropped rather than done by me

Why it works: “Only-I-can-do-this” is the most important filter for solo operators and founders. Many tasks on a long list don’t actually require you. The “two hardest calls” question forces Claude to surface its reasoning rather than just presenting a list.

Prompt 5: The Weekly Retrospective

Use this: Friday afternoon, 10–15 minutes.

End-of-week retrospective. Today is [date].

I planned to do this week:
[paste Monday's plan]

What actually happened:
[honest account — what got done, what didn't, what surprised you]

Items that got pushed (things I planned but didn't do): [list]
Things that appeared and ate time that weren't in the plan: [list]

Diagnose my planning from this week. Be specific:
1. Where were my time estimates wrong and by how much?
2. Is there a pattern in what got pushed (e.g., same type of task, same time of day)?
3. One structural change to how I should set up next week's plan.

Don't just tell me to be more realistic. Tell me exactly where the plan failed.

Why it works: The instruction “don’t just tell me to be more realistic” prevents the vague encouragement mode that retrospective conversations often default to. The three specific questions direct Claude’s analysis toward things you can actually act on next Monday.

How to Get More from These Prompts

Each prompt improves when run inside a Claude Project where your role, goals, and constraints are pre-loaded. Without that context, Claude has to infer your situation from what’s in the prompt.

With a Project configured, Claude already knows your quarterly goals, your standing meetings, and the failure patterns you’ve previously identified. The prompts then produce more specific, more accurate output with less manual input.

Your action: Copy Prompt 1 right now. Fill in your tasks for today, paste it into Claude, and run it. That’s the fastest way to see whether this approach works for your situation.

Related: The Complete Guide to Planning with Claude AI · How to Plan with Claude AI Step by Step · Why Claude “Refuses” to Plan Your Day · 5 AI Prompts for Goal Setting

Tags: Claude prompts for planning, AI planning prompts, Claude daily planning, weekly plan Claude, Claude project decomposition