Guide · Agent skill
How to use the Todo Write Planning agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The todo-write skill is a small instruction set for an AI agent. It tells the agent to write a numbered, step-by-step plan using its built-in plan or task tool before it starts creating files, then tick items off as each one finishes. This guide is for people who have never used agent skills as well as people already working in OpenDesign. By the end you will know what the skill changes, how to install it, how to ask for it, and where its limits are.
Todo Write Planningnexu-io/open-design · Apache-2.0 Download the skill Quick start
- 1Download the ZIP. Download the todo-write skill ZIP from its page on this directory.
- 2Unzip the skill. Extract the archive so the open-design.json and SKILL.md files sit in a todo-write folder.
- 3Choose the install location. Place the folder in ~/.claude/skills/ for personal use, or in a project's .claude/skills/ folder for one project.
- 4Or upload it in the Claude apps. Under Settings → Capabilities → Skills, upload the skill if you use the Claude apps instead.
- 5Confirm your session has a plan tool. Make sure the agent session exposes a plan or todo tool, since the skill does not name one.
- 6Ask for a plan before generation. Give the agent a multi-step task and ask it to plan first and tick items off as it works.
What the todo-write skill makes the agent do
Before writing any artifact files, the agent commits to a numbered plan using its plan tool. The plan works as an audit trail: as the work continues, the agent either ticks items off or rewrites the plan.
The skill teaches the agent four habits: keep each todo atomic with one verb per todo, reorder freely as the picture gets clearer, mark a todo complete only after the matching artifact has been written, and surface blockers as todos instead of silently skipping them.
- Plan first, then generate, for non-trivial workflows
- One verb per todo item
- Mark items complete only after the related artifact exists
- Record blockers as visible todos
What this skill is and is not
This skill is a planning habit, not a new capability. It does not add any tools or external connections. It works by placing the instruction to make a plan first into the agent's system prompt.
Because it is only an instruction, the agent can still deviate from the plan. If you need a guarantee that every step is followed, this skill alone does not provide one.
Install the todo-write skill
The skill is distributed as a ZIP from the skill's page on this directory. Unzip it into one of the locations below. The package contains two files: open-design.json and SKILL.md.
You can also upload the skill in the Claude apps under Settings → Capabilities → Skills.
- Personal install: unzip into ~/.claude/skills/
- Project install: unzip into the project's .claude/skills/ folder
Trigger the skill with natural requests
The source does not list specific trigger phrases, so the simplest approach is to describe a longer task and ask for a plan first. The skill is meant for non-trivial workflows, so give the agent a multi-step job rather than a single quick question.
Ask the agent to plan before it generates anything, and to show progress as it works. The plan appears as a live Todos card when the OpenDesign host renders it.
- Plan the steps first, then build the landing page files
- Make a checklist for this multi-file design project and tick items off as you finish them
- Before creating any files, write a numbered plan and tell me why you ordered the steps this way
What the skill needs to work
The skill needs an agent session that exposes a plan or todo tool. The skill does not name a specific tool, so use whichever plan tool your session actually provides. The OpenDesign host renders every such tool as the same Todos card, which is optional and only affects the visual display.
The skill runs no scripts and uses no network access.
- An agent session with a plan or todo tool
- OpenDesign host to show the Todos card (optional)
- Runs scripts: no
- Uses network: no
Safety notes
Because the skill runs no scripts and uses no network, its direct risk is low. It only changes the instructions the agent follows about planning. Review the included SKILL.md file if you want to see exactly what is being added to the agent's prompt.
Keep in mind that the plan is a record of intent, not proof of what was done. Check the actual artifacts the agent produced.
Tips for getting good plans
Keep requests substantial enough to justify a plan. Ask the agent to reorder the plan when you change direction, since the skill allows reordering as the picture sharpens. When a step is blocked, look for it as a todo item rather than assuming it was skipped.
- Use it for multi-step design or build tasks
- Review the plan before generation starts when the work is expensive to redo
- Expect the plan to change as the work progresses
What people use it for
- Planning a multi-page design project before any HTML or asset files are created
- Building a longer AI-generated project with a visible checklist of progress
- Recording blockers as todos so they are not silently skipped
- Reviewing the agent's intended order of steps before it starts generating
Frequently asked questions
How do I install the todo-write skill?
Download the ZIP from the skill's page and unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder. In the Claude apps you can instead upload it under Settings → Capabilities → Skills.
What does the todo-write skill require?
It needs an agent session that exposes a plan or todo tool. The skill does not name a specific tool. The OpenDesign host is optional and only renders the Todos card.
Does the todo-write skill run scripts or use the internet?
No. The skill runs no scripts and uses no network access. It only adds an instruction to plan before generating.
Is the todo-write skill free?
It is licensed under Apache-2.0. Check the skill's page on this directory for any other details about its licence.
Will the agent always follow the plan?
Not guaranteed. The skill is only an instruction in the system prompt, so the agent can still deviate from the plan. Check the final artifacts to confirm what was done.
When should I not use todo-write?
Skip it for quick one-line questions or small edits, where a plan adds overhead without benefit. It is meant for multi-step design or build tasks.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use todo-write if you work in OpenDesign or a similar setup and want longer design or build tasks planned and tracked in a visible checklist. Pick something else if you need a guaranteed workflow, since this skill only asks the agent to plan, and skip it for quick, small requests where a plan adds little.
Download the skill