Skip to main content

Developer Tool CLI & IDE Plugin UI

Design developer tool UIs with command palettes, keyboard-first navigation, and information-dense layouts. Built for engineers who judge tools in 30 seconds.

Fill in the placeholders

Edit the values, then copy your finished prompt.

Your Prompt
prompt.txt

                                

What this prompt does

This prompt generates a complete, opinionated UI specification for developer tools — covering the command palette, workspace layout, keyboard shortcut system, output panel, settings interface, and status bar as discrete, implementation-ready sections. It works because it encodes the actual conventions developers have internalized from tools like VS Code, Zed, and Warp: monospace-first typography, dark charcoal backgrounds, sub-100ms feedback, and syntax-highlighting color palettes applied to UI accents — not just code.

The anti-pattern list is the real differentiator. By explicitly instructing the model to avoid rounded cards, slow animations, and light-mode-only designs, the prompt prevents the default "pretty SaaS" aesthetic that makes developer tools feel wrong the moment you open them. The [framework] variable drives the final section, which generates an implementation of the command palette and workspace layout in your chosen stack — so you get runnable component code for those two critical sections, not just a prose spec.

The [integration_context] variable is where you describe where the tool lives: a VS Code extension sidebar, a standalone Electron window, a JetBrains plugin panel, or a browser DevTools extension. This context changes every interaction assumption the model makes — a sidebar panel gets a condensed single-column layout with collapsed panels; a standalone window gets full split-pane workspace treatment.

When to use it

  • You are building a CLI companion app (log viewer, deployment dashboard, database explorer) and need a coherent visual system before writing a line of frontend code.
  • Your team is designing a VS Code extension or JetBrains plugin and wants to align on interaction patterns before splitting into parallel work.
  • You are pitching a developer tool to stakeholders and need a realistic, high-fidelity mockup specification rather than generic Figma frames.
  • You are evaluating whether a tool's UX matches developer expectations before a public beta.
  • You inherited a developer tool with inconsistent UI and need a reference spec to refactor toward.

Example output

For [tool_type]: "Kubernetes log aggregator", [developer_persona]: "Platform engineer running 20+ services", [core_workflow]: "Filtering and tailing logs across namespaces in real time", [integration_context]: "Standalone Electron window", [framework]: "React + xterm.js":

Command Palette (Cmd+K):
  > namespace:production filter:error last:5m
  > pod:api-gateway tail --follow
  Recent: [namespace:staging filter:warn] [pod:worker-3 grep "OOM"]
  Categories: Namespaces | Pods | Filters | Views | Settings

Workspace: split-pane — namespace tree (left 240px) | live log stream (main) | parsed JSON inspector (right 320px, toggleable)

Keyboard shortcuts:
  Ctrl+Shift+F  — focus filter input
  Ctrl+\         — toggle JSON inspector
  G then G       — scroll to bottom of log stream (chord)
  Ctrl+K         — clear current stream

Output panel: xterm.js instance, ANSI color passthrough, auto-scroll locked unless user scrolls up (banner: "▲ Paused — press G to resume")

Settings: JSON editor view default, GUI toggle in top-right; schema-validated, errors inline

Status bar: connected to cluster prod-us-east | 3 streams active | latency 12ms | branch: main

(Truncated — the full prompt output also includes the React + xterm.js implementation of the command palette and workspace layout.)

Pro tips

  • Set [developer_persona] with precision: "backend engineer who lives in the terminal" generates a denser, more keyboard-centric output than a vague "developer." The persona shapes every section — shortcut density, information hierarchy, even the default panel layout.
  • The [workspace_layout] variable is the highest-leverage input. Describe the layout structurally — "three-column: file tree | diff view | AI suggestions panel" — not conceptually like "clean and organized." The model derives keyboard navigation and panel resize behavior from the structural description.
  • The [integration_context] variable is not cosmetic. "VS Code extension sidebar" versus "standalone Electron window" produces fundamentally different workspace layouts and keyboard shortcut systems — a sidebar gets constrained width with collapsed sections, a standalone window gets full multi-pane treatment. Fill this in before anything else.
  • The [framework] slot accepts both frontend stacks (React, Svelte, Tauri) and terminal UI libraries (Ink, Bubble Tea, Textual). The command palette and workspace implementation adapts meaningfully to each — Ink gives you a terminal-rendered palette with keyboard event handlers; React gives you a DOM-based cmdk-style component.
  • If your tool has a light-mode requirement (enterprise compliance, accessibility), explicitly override the dark-mode default in your prompt. The anti-pattern instruction will otherwise suppress light-mode output entirely, which is intentional for most use cases but will silently omit light theme variants if you need them.

Frequently Asked Questions

Can I use this prompt for a terminal-only tool with no graphical UI?
Yes. Set [framework] to a TUI library like Bubble Tea (Go), Textual (Python), or Ink (Node.js), and set [integration_context] to 'terminal, no GUI.' The command palette section generates keyboard-driven navigation using terminal-native patterns — no DOM elements, no mouse dependency. The output panel section maps directly to stdout with ANSI color support, which is already baked into the prompt's design rules. The settings interface section will default to a TOML or JSON config file approach rather than a GUI toggle.
The prompt generates a specification — how do I get actual component code from it?
The [framework] variable drives code generation for the command palette and workspace layout sections. When you specify 'React + Radix UI' or 'Tauri + SolidJS', the final section of the prompt explicitly asks for an implementation in that stack — you get component code for those two sections, not just prose. The remaining five sections (keyboard shortcuts, output panel, settings, status bar, and the full workspace spec) are implementation-ready descriptions, not code. To convert those, take each section and pass it back to the model with 'now implement this as [framework] code' as a targeted follow-up.
How specific does [core_workflow] need to be?
Specific enough to describe the main loop a user repeats dozens of times per day. 'Writing and running database migrations against a staging environment' works. 'Using the tool' does not. The workspace layout, keyboard shortcuts, and output panel the model generates are all derived from the core workflow — the keyboard shortcut chord system for a log aggregator (scroll, filter, clear) will be completely different from one designed for a migration runner (apply, rollback, diff). Vague input produces a generic output that could describe any developer tool and satisfies none of them.
Engr Mejba Ahmed

Need this built for real?

Engr Mejba Ahmed

AI Developer · Software Engineer

I'm Mejba — I design and ship production AI systems, automations, and full-stack apps. If you want this turned into a working solution for your team, let's talk.

More in Tech & SaaS UI Prompts

Engr Mejba Ahmed

Engr Mejba Ahmed

Claude Code Expert · Online

👋

Hey there!

Quick Actions

WhatsApp Instant reply

Chat on WhatsApp

+880 1723 741224 · Instant reply

Popular Questions

Engr Mejba Ahmed is connected
Engr Mejba Ahmed is typing...
Engr Mejba Ahmed avatar

✉ Want me to follow up? Drop your email

Engr Mejba Ahmed avatar

📞 Connect Directly

Choose how you'd like to reach me

WhatsApp

+880 1723 741224

Email

[email protected]

✓ Details sent! I'll get back to you shortly.

Powered by OpenAI

335+

Blog Posts

25

AI Courses

63

Projects

Services & Expertise

Pricing & Process

Learning & Resources

Connect & Support