microinteractions
Design the small details -- triggers, rules, feedback, loops and modes -- that separate good products from great ones
Install / Use
npx skills add wondelai/skills --skill microinteractionsInstalls into whichever agent you are using.
SKILL.md
Installable skill definition
Quality Score
Category
Development & EngineeringSupported Platforms
Our assessment of microinteractions
microinteractions scores 84/100 on our quality scale, 2882nd of 4,700 Development & Engineering skills we index.
Its SKILL.md is 15 KB long, well organised into 14 sections and no code examples: a thorough specification that gives an agent plenty to work with.
With 2,248 GitHub stars, it is one of the more widely adopted skills in the catalogue.
Maintenance, license and trust
- The repository was last updated about 2 months ago, so microinteractions is actively maintained.
- Our last check on 2026-11-05 found the source still online.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-10-01. Automated pattern scan on 2026-10-01. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
microinteractions compared with similar skills
All 4 of these similar skills score higher than microinteractions; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| microinteractions (this skill)by wondelai | 84 | 2.2k | 2mo ago | SKILL.md |
| taste-skillby Leonxlnx | 100 | 116.8k | 8d ago | CLAUDE.md |
| ai-job-searchby ai-job-search-dev | 100 | 72.1k | today | CLAUDE.md |
| claude-howtoby luongnv89 | 100 | 47.3k | 2d ago | CLAUDE.md |
| Claude-Code-Game-Studiosby Donchitos | 100 | 39.7k | 7d ago | CLAUDE.md |
Frequently asked questions
- How do I install microinteractions?
- Run
npx skills add wondelai/skills --skill microinteractions. The install tabs above show the steps for each supported agent. - Which AI agents does microinteractions work with?
- It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
- Is microinteractions safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is microinteractions still maintained?
- The repository was last updated about 2 months ago, so microinteractions is actively maintained.
Skill content
View source on GitHubname: microinteractions description: 'Design the small details -- triggers, rules, feedback, loops and modes -- that separate good products from great ones. Use when the user mentions "microinteraction", "button feedback", "loading state", "toggle design", "animation detail", "state transitions", "input feedback", "the interface feels dead", "make the UI feel responsive", or "add polish to interactions". Also trigger when designing form-validation responses, progress indicators, confirmation dialogs, or any element where the user expects immediate feedback. Covers trigger design, state rules, feedback mechanisms, and progressive loops. For overall UI polish, see refactoring-ui. For affordance design, see design-everyday-things.' license: MIT metadata: author: wondelai version: "1.4.1"
Microinteractions Framework
Design the tiny, contained product moments users touch every day -- toggles, password fields, loading indicators, pull-to-refresh, like buttons. Based on Dan Saffer's four-part structure (Trigger, Rules, Feedback, Loops & Modes), this framework turns invisible details into the polish that separates forgettable products from beloved ones.
Core Principle
The difference between a product you tolerate and a product you love is almost always in the microinteractions. A microinteraction is a contained moment built around a single use case -- changing a setting, syncing data, picking a password -- so small that users rarely think about it consciously, but they feel it. Every microinteraction follows the same four-part structure: a Trigger initiates it, Rules determine what happens, Feedback shows what is happening, and Loops & Modes define its long-term behavior.
Scoring
Goal: 10/10. Score by how many of the 8 Quick Diagnostic rows the microinteraction passes — score = round(passed / 8 × 10), then read the band:
- 9-10 = passes all 8 rows: deliberate discoverable trigger with visible states, simple predictable rules, sub-100ms feedback scaled to event significance, evolves over time, mode-free or mode-visible, learnable without help.
- 5-6 = 4-5 rows pass: it works but has a generic feel -- e.g. feedback exists but is uniform, or the trigger lacks distinct states.
- <=3 = 2 or fewer rows pass: missing feedback, invisible triggers, or hidden modes that break trust.
Always state the current score, which diagnostic rows failed, and the specific fix for each.
The Microinteraction Structure
Six areas of focus for designing world-class microinteractions. See references/case-studies.md when you want a full four-part breakdown of a real pattern -- form submission, toggle/switch, pull-to-refresh, loading states, and notifications, each from first use through edge cases.
1. Triggers
Core concept: The trigger initiates a microinteraction -- manual (tap, click, swipe, voice command) or system-initiated (time, location, incoming data, error state). It is the front door of every microinteraction.
Why it works: A trigger's prominence and labeling set the user's expectation before they act -- a button that reads "Delete" in red signals an irreversible, high-stakes outcome, so feedback that follows feels predictable rather than surprising.
Key insights:
- A trigger must communicate three things: that it exists, what it does, and what state it is in
- Match trigger prominence to action importance -- high-stakes actions need prominent triggers
- Pair invisible triggers (gestures, shake, proximity) with a visible alternative for discoverability
- Make trigger states -- default, hover, active, disabled, loading -- visually distinct
Product applications:
| Context | Application | Example | |---------|-------------|---------| | Toggle controls | Manual trigger with binary state | iOS Wi-Fi switch: tap to toggle, position shows state | | Pull-to-refresh | Hidden gesture with visible affordance | Pull past threshold triggers refresh animation | | System alerts | System trigger on condition met | Low battery notification at 20% threshold |
Ethical boundary: Never hide critical triggers behind gestures or invisible interactions without a visible fallback.
See: references/trigger-design.md for trigger affordances, states, placement, and reducing trigger complexity.
2. Rules
Core concept: Rules define what happens once a microinteraction is triggered -- the sequence of events, constraints, processing, and ending. Users never see rules directly, but they feel when rules are wrong.
Why it works: Rules create the mental model users build about how the interaction works. Consistent rules that match expectations feel natural; violations -- a toggle that does not toggle, a slider that jumps in value -- destroy trust.
Key insights:
- Define the goal of the microinteraction first, then derive rules from it
- Match existing mental models and platform conventions
- Constrain inputs to prevent errors: limit character counts, set value ranges, enforce formats
- Handle edge cases explicitly: zero, maximum, repeated triggers, interruption
Product applications:
| Context | Application | Example | |---------|-------------|---------| | Password strength | Rules evaluate input in real-time | Meter updates as user types; color shifts red to green | | Character counter | Rule constrains and shows remaining | Twitter/X: counter decreases, turns red at limit | | Undo action | Rule sets time window for reversal | Gmail "Undo send" available for 30 seconds |
Ethical boundary: Keep rules transparent and predictable -- never hide rules that manipulate behavior, such as making unsubscribe harder than subscribe.
See: references/rules-and-state.md for state management, constraints, error states, and edge cases.
3. Feedback
Core concept: Feedback communicates the rules to the user, answering "What is happening right now?" -- visually (color, animation, movement), aurally (clicks, chimes), or haptically (vibration). Show only what matters: minimal, meaningful, contextual.
Why it works: Without feedback, users cannot tell if their action registered or the system is working, so they retap, abandon, or distrust the result. Feedback is what converts an invisible system state into a perceived response.
Key insights:
- Feedback must be immediate -- under 100ms for direct manipulation
- Use the least noticeable feedback that still communicates, and prefer animating existing elements (the button itself, not a separate toast)
- Scale feedback to event significance: small action = small feedback, big result = big feedback
- Visual feedback is primary; audio and haptic are supplementary, never the only channel
- Progress indicators reduce perceived wait time even when actual time is unchanged
Product applications:
| Context | Application | Example | |---------|-------------|---------| | Button press | Visual state change on click | Button depresses, color shifts, text becomes "Saving..." | | Form validation | Inline feedback as user types | Green checkmark next to valid email field | | Error state | Contextual error near the source | Red border on field + "Password must be 8+ characters" |
Ethical boundary: Keep feedback honest -- no fake progress bars, manipulative countdowns, or deceptive completion percentages.
See: references/feedback-patterns.md for feedback channels, timing, and preventing overload.
4. Loops and Modes
Core concept: Loops are the meta-rules over time -- does the interaction change after the 100th use, expire, adapt? Modes are forks in the rules where the same control temporarily behaves differently (edit mode vs. view mode).
Why it works: Thoughtful loops let microinteractions mature gracefully -- reducing friction for power users while staying discoverable for new ones. Modes, used sparingly, let one control serve multiple purposes without cluttering the interface.
Key insights:
- Open loops continue until explicitly stopped (a repeating alarm); closed loops run once and end (a timer)
- Long loops change the interaction over time: first use shows a tooltip; the 50th does not
- Progressive reduction: strip away scaffolding as users demonstrate mastery
- Modes are dangerous -- they violate "same action, same result"; minimize them and make the current mode highly visible (Caps Lock indicator, edit banner)
Product applications:
| Context | Application | Example | |---------|-------------|---------| | Onboarding tooltips | Long loop removes hints after N uses | First 3 sessions show "Swipe to archive"; then stop | | Alarm clock | Open loop repeats until disabled | Fires every weekday at 7am until toggled off | | Text editing | Mode: view vs. edit | Banner reads "Editing" with a "Done" button to exit |
Ethical boundary: Loops should benefit the user, not the business -- never adapt loops to ramp up notifications or make opt-outs progressively harder.
See: references/loops-modes.md for long loops, mode errors, and progressive complexity.
5. Signature Moments
Core concept: A signature moment is a microinteraction so distinctive it becomes part of the product's identity -- the Facebook Like, slide-to-unlock, Slack's loading messages. Every product should have one or two; not every interaction should be one.
Why it works: Signature moments create emotional memory and make products feel crafted rather than assembled. They are what users demonstrate first when describing your product to others.
Key insights:
- Put signature moments on frequent, visible actions -- not buried settings
- Functional first, delightful second -- never sacrifice usability for novelty
- Animation, sound, and copy are the three most common tools
- Align with brand personality: playful brands get playful moments
- Apply the removal test: if users would not miss it, it is decoration, not signature
Product applications:
| Context | Application | Example | |---------|-------------|---------| | Social reaction | Animated response to engagement | Facebook Like: thumbs-up animates with particles | | Loading state | Branded waiting experience | Slack: rotating quotes during load | | Completion | Celebratory confirmation | Stripe payment: animated checkmark with confetti |
Ethical boundary: Never block input or the next step behind a non-skippable celebration animation -- let the user tap through the confetti to proceed.
See: references/signature-moments.md for when to invest and making mundane interactions delightful.
6. Reducing and Simplifying
Core concept: The best microinteraction is barely noticed because it is so simple and fast. Reduce (fewer options, steps, decisions), then simplify what remains until it feels effortless.
Why it works: Every option, field, and decision adds cognitive load -- users do not want to configure a toggle, they want it to work. The most elegant microinteractions have zero configuration, one action, and immediate results.
Key insights:
- If a microinteraction needs instructions, it is too complex
- Remove options with smart defaults -- pick the best choice and commit to it
- Collapse multi-step interactions into a single action where possible
- Use progressive disclosure: show simple first, reveal complexity only on request
- Keep rule count proportional to frequency of use: common actions need few rules
Product applications:
| Context | Application | Example | |---------|-------------|---------| | Smart defaults | Eliminate configuration | Camera app opens in photo mode, not settings | | Single action | One tap replaces multi-step flow | Double-tap to like instead of menu + select reaction | | Anticipatory design | Predict and pre
Truncated for display — read the full file on GitHub.
Related Skills
taste-skill
116.8kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
ai-job-search
72.1kThe job search that runs on your machine. AI job application framework built on Claude Code: evaluate postings, tailor CVs, write cover letters, prep interviews. Fork it and own it.
claude-howto
47.3kA visual, example-driven guide to Claude Code — from basic concepts to advanced agents, with copy-paste templates that bring immediate value.
Claude-Code-Game-Studios
39.7kTurn Claude Code into a full game dev studio — 49 AI agents, 72 workflow skills, and a complete coordination system mirroring real studio hierarchy.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
