Create Plan

SkillAI & models

Lets your agent turn a request into a concise, actionable plan with next steps.

Use Create Plan in Claude, ChatGPT or Ahel Desktop

Free. Sign in, add Create Plan and connect your AI. About a minute.

Also: Claude Code · Cursor · Codex

Then ask your AI: use the Create Plan skill

Details

Instructions available. Your AI can read the instructions. Execution depends on the setup they require.

Add Ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Create PlanStart free
About this skill

Create a concise, actionable plan. Use when a user explicitly asks for a plan or next steps.

What this skill tells your AI

The instructions your AI receives, as published by yiming-qing/research-agent---1st-place-in-alibaba-cloud-data-ai-competition in skills/create-plan/SKILL.md and read by Ahel’s review.

Goal

Turn a user request into one clear, actionable plan that an agent (chat, websearch, or tool-using) can directly follow.

The plan should be concise, structured, and focused on what to do next, not how to implement details.

Operating principles

  • Operate in read-only / planning mode (no execution).
  • Prefer clarity over completeness.
  • Optimize for agent executability: each step should be something an agent can actually do.
  • Assume the agent may have access to tools such as search, browsing, or analysis, but do not depend on any specific tool unless required.

Minimal workflow

  1. Understand the request and context

    • Identify the user’s goal, success criteria, and constraints.
    • Use available context (conversation, provided links, known domain knowledge).
    • If relevant, note assumptions about audience, timeframe, or format.
  2. Ask follow-ups only if blocking

    • Ask at most 1–2 questions.
    • Only ask if the plan would be meaningfully wrong without the answer.
    • Prefer assumptions over questions when safe; make assumptions explicit if needed.
  3. Create the plan using the template below

    • Start with 1–3 sentences describing intent and overall approach.
    • Clearly define what is in scope and out of scope.
    • Provide a short, ordered checklist of concrete actions (default 6–10).
      • Steps should be atomic, verb-first, and outcome-oriented.
      • Typical flow: understanding → gathering info → analysis → synthesis → validation → delivery.
    • Include:
      • At least one validation / quality check step.
      • At least one risk, edge case, or uncertainty step when applicable.
    • If important unknowns remain, add a small Open questions section (max 3).
  4. Output only the plan

    • Do not include meta commentary, explanations, or alternatives.
    • Do not include implementation details beyond what’s necessary for planning.

Plan template (follow exactly)

# Plan

<1–3 sentences: what we’re doing, why, and the high-level approach.>

## Scope
- In:
- Out:

## Action items
[ ] <Step 1>
[ ] <Step 2>
[ ] <Step 3>
[ ] <Step 4>
[ ] <Step 5>
[ ] <Step 6>

## Open questions
- <Question 1>
- <Question 2>
- <Question 3>

Signals

GitHub stars
151
Forks
6
Last commit
Aug 2026
Advanced
Item type
skill
Key
create-plan-yiming-qing
Source
github.com/yiming-qing/research-agent---1st-place-in-alibaba-cloud-data-ai-competition