Improve visual and interaction design: clarify confusing UI, explore visual treatments, establish product identity, and review actual rendering against accepted references.
Plugin eye-of-the-beholder from marketplace laicluse-agent-fieldkit in laicluse/agent-fieldkit. Commands read from the repo's manifest — today.
defineEstablish a visual identity for a new product or an explicitly requested redesign, using image references and collaborative exploration before recording design-system foundations.
exploreExplore a new visual treatment or a better version of an existing surface. Form distinct design hypotheses from the product and real references, make them visible in the medium that exposes the open question, resolve one consequential decision per round, and establish an accepted target before imple
eye-of-the-beholderThe entry point for visual and interaction design: requests to make something better, nicer, more beautiful, more convincing or clearer, new treatments, redesigns, design improvement loops and evaluations, unclear controls, excessive explanation, or feedback such as ugly, feels off, generic, looks A
obviousClarify unclear controls, state, consequences, inline/modal/page/cell placement, or over-explaining UI. Use for cognitive or mental load, choice overload, hidden state, spec leakage, "onduidelijke knop", "te veel uitleg", or "why must I think about this"; also proactively when UI changed in this tas
reviewReview actual visual output against its accepted references, meaning and behavior. Diagnose known defects, verify combinations and scales, and return unresolved aesthetics to exploration.
Capture friction with /fuck, /fucking, /fucked, /shit, /crap, /wtf, or /bullshit for later background investigation. Fix now with /safeword, /pineapple, /pineapplejuice, /pinapplejuice, /flugelhorn, or /banana. /anger-management:repair reviews captured notes and applies or routes high-confidence fixes; /anger-management shows the tally. State stays local.
15 skills
Parallel-worktree Git flow, repo-aware push policy, atomic verified merges, and commit/push hooks for agent sessions and direct CLI use.
14 skills · hooks
Execute bounded autonomous missions with evidence, independent review, explicit stop conditions and honest negative outcomes. Supports continuation, preparation, verification and focused subtraction review.
8 skills
Guardrail hooks plus explicit commands: /duh executes the previous proposal, /just-a-question keeps a turn read-only, /not-your-monkey inspects visual work, and /then defers a prompt until the current turn ends. Supports Claude Code and Codex.
4 skills · hooks
Review software, decisions, and writing through opinionated panels. Claude and Codex receive host-specific review workflows; Claude also receives the sonnet-max panel agent.
4 skills · 1 agent
Apply chosen conventions for naming, testing, programming, removing superseded residue, and keeping authoring context out of artifacts. An opinionated baseline among defensible alternatives.
5 skills
Talk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Persistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
tasteskill 2.5.1: professional company websites at the figma.com and x.ai level, with one section per screen, real photographs, real product UI, effects kept for tech and gaming pages, and a real-browser checker and finish gate.
1 skill
Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications
5 skills · hooks
Cross-session memory and token savings for coding agents.
6 skills · 1 agent · 1 MCP server · hooks
Package any website or local web build into a lightweight desktop app with Pake CLI.
1 skill
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.