Guide · Agent skill
How to use the Daily UI Inspiration Capture agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The daily-ui-inspiration-capture skill is an agent skill that tells an AI agent how to run a repeatable daily routine: gather five recent landing-page designs, save screenshots and motion frames for each, and package them into a dated article folder with a written prompt for each one. This guide explains what the skill makes the agent do, how to install it, how to ask for it in plain language, and what it needs before you start. By the end you will know whether it fits your project and how to run it safely.
Daily UI Inspiration CaptureMengTo/Skills · MIT Download the skill Quick start
- 1Download the skill ZIP. Get the ZIP of daily-ui-inspiration-capture from the page where you found it.
- 2Unzip into a skills folder. Unzip into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project.
- 3Or upload it in the Claude apps. Open Settings, go to Capabilities, then Skills, and upload the skill.
- 4Check the requirements. Confirm Node.js is installed, your content project has an articles/ folder and is a git repository, and you have the browser and capture tools the workflow expects.
- 5Make a natural request. Ask the agent to run or refresh today's UI inspiration capture and name the project if you have more than one.
- 6Review the bundle. Check the dated articles folder, content.md and manifest.json, then run the duplicate checker on the manifest.
What the skill makes the agent do
An agent skill is a folder of instructions that an AI agent reads when your request matches the skill's description. This one covers a daily UI inspiration routine. The agent collects five final references, usually live landing pages such as Framer Marketplace sites, and writes them into a folder named by date under articles/ in your content project.
For each reference the skill asks for a still image, a full-page scroll screenshot cut into section crops from hero to footer, and motion frames from a short screen recording when video capture works. It also asks for a long prompt that another AI builder could use to recreate the feel of the page. Before choosing finalists, the agent loads earlier manifest.json files and rejects repeats by title, source URL or image URL. The output ends with a content.md file and a manifest.json file.
- Five references per bundle, with a fallback reason documented when video or live capture is blocked
- Full-page screenshot plus contiguous section crops for each live website
- Motion frames when video succeeds, with at least four per item
- A duplicate check against previous captures before committing
Install the skill with a ZIP download or the Claude apps
There are two routes. For Claude Code, download the ZIP of the skill from the page where you found it, then unzip it into your personal skills folder at ~/.claude/skills/. To scope it to one project instead, unzip it into that project's .claude/skills/ folder. Restart or start a new session so the agent can see the new skill.
In the Claude apps, open Settings, go to Capabilities, then Skills, and upload the skill there. The skill folder is what gets loaded, so keep the files together and do not separate SKILL.md from the rest of the folder.
- Personal install: ~/.claude/skills/ (available across projects)
- Project install: .claude/skills/ inside the project folder
- Claude apps: Settings, then Capabilities, then Skills, then upload
Ask for the skill with natural requests
You do not need a special command. The agent picks up the skill when your request matches its description. The description mentions running, refreshing, packaging or validating dated UI inspiration bundles, Framer or Dribbble landing-page inspiration, motion-study screenshots or videos, AI-builder prompts, duplicate checking, and turning a project runbook into a repeatable workflow.
Name the project if you have more than one. If you do not, the skill says the agent will use the current workspace and report which directory it used. Including the words articles, UI inspiration or duplicate check in your request makes the match more reliable.
- Run today's UI inspiration capture for this content project.
- Refresh the dated UI inspiration bundle and avoid anything in earlier captures.
- Validate the current ui-inspiration-capture bundle and run the duplicate checker.
- Turn our existing runbook into a repeatable daily landing-page inspiration workflow.
Run the verification step
Before committing, the skill asks the agent to run the duplicate checker against the new manifest. The command takes the path to that bundle's manifest.json. Replace the placeholder with the real dated folder name the agent created.
The checks the skill lists include duplicateCount equal to 0, exactly five items in manifest.json, at least five still images embedded in content.md, and every referenced still, video and motion frame existing on disk. Each live website should have a full-page screenshot and section crops, and each item should have either a local video or a stated fallback reason.
- Confirm the README.md file is absent, since the skill says not to create one.
- Articles folders are often git-ignored, so the skill expects the agent to use git add -f when staging them.
- If articles/ARTICLES_INDEX.md is dirty or conflicted, the skill says to leave it alone and mention that in the summary.
node scripts/check-ui-inspiration-duplicates.mjs articles/YYYY-MM-DD-ui-inspiration-capture/manifest.jsonKnow what the skill needs and its safety limits
The skill expects the Codex in-app browser for any browser work and says not to use Chrome. It needs Node.js to run the duplicate checker script. It also needs a content project that has an articles/ folder and is a git repository. The skill files do not name the screen capture or video frame tools the workflow relies on, so you must make sure your environment provides them.
The skill runs scripts and uses the network, so read the scripts in your project and the sites the agent visits before you run it. The skill does not address licensing, so the agent pushes reuse through a prompt rather than reproducing designs verbatim. Do not use it on sites behind logins or paywalls. If a source blocks inspection, the skill asks the agent to use the best available evidence or a local asset and explain why.
- Requires the Codex in-app browser; the skill says not to use Chrome
- Requires Node.js for scripts/check-ui-inspiration-duplicates.mjs
- Requires a content project with an articles/ folder and a git repository
- Runs scripts and uses the network
- Does not cover licensing, so review the rights to any design before reuse
Tips for getting useful bundles
The skill is strongest when your project already has a runbook or a recent bundle. The agent reads the latest one or two content.md and manifest.json files to match the current shape, so keep those files in place. Clear, specific briefs also help. Say which kind of page you want, such as a SaaS hero or a product site, and rotate search topics if the feed feels stale.
Favor references with a clear first viewport, visible sections and some motion. The skill itself says to avoid vague atmospheric thumbnails. Treat the prompts as starting points for your own build, not copies of someone else's page.
- Keep AGENTS.md and scripts/ui-inspiration-capture-guidelines.md in the project if you have them; the skill reads them when present.
- Rotate searches such as website hero, saas landing page, product website or recent product design when results repeat.
- Ask the agent to say which directory it used when you have several projects.
What people use it for
- A designer collects five annotated landing-page references each morning for a moodboard.
- A content creator publishes a daily UI inspiration article with the same structure every day.
- A front-end team turns its existing capture runbook into an agent-driven routine.
- An agent validates an older bundle for duplicates and missing files before it is published.
Frequently asked questions
How do I install the daily UI inspiration capture skill?
Download the ZIP of the skill from the page where you found it and unzip it into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project. In the Claude apps, go to Settings, then Capabilities, then Skills, and upload it there.
Is the daily-ui-inspiration-capture skill free?
The source material for this page does not state a price or licence terms for use beyond the repository's MIT licence. The skill's own files do not address licensing for the designs it captures, so check the rights to any design before reusing it.
What do I need to run the skill?
You need the Codex in-app browser for browser work, Node.js to run the duplicate checker script, and a content project with an articles/ folder and a git repository. The screen capture and video frame tools are referenced but not specified in the skill files.
Does the skill work outside a project with an articles folder?
The skill is built around a specific layout with articles/ and scripts/ folders. Without that layout the workflow is long and may not fit, so it is better suited to a content project that already follows it.
Is it safe to run this skill?
The skill runs scripts and uses the network. Read the scripts in your project and review the sites the agent visits before running it. Avoid sites behind logins or paywalls, which the skill is not meant to capture.
What if the agent cannot capture a site?
The skill asks the agent to use the best available browser evidence or a local asset fallback and to document the reason. It also says to prefer stronger Framer or direct-page candidates instead of forcing brittle workarounds.
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
This skill suits designers, front-end developers and content creators who want a steady, consistently structured stream of annotated landing-page inspiration and already work in a project with an articles/ folder. It is a poor fit if you only want a few screenshots, need to capture logged-in or paywalled sites, or plan to reuse designs without checking their rights.
Download the skill