Gentleman-Programming/gentle-shell / skills/tui-design
gentle-ai-tui-design
VerifiedTrigger: terminal UI design, TUI redesign, Gentle Shell look, IDE in a terminal, sidebar/status bar polish, critique TUI. Refine terminal UIs with visual proof.
| File | Kind | Size |
|---|---|---|
| NOTICE.md | markdown | 403 B |
| SKILL.md | skill md | 3.3 KB |
| references/commands.md | markdown | 3.1 KB |
| references/critique.md | markdown | 2.1 KB |
| references/design-doc.md | markdown | 1.4 KB |
| references/terminal-craft.md | markdown | 3.9 KB |
| references/visual-loop.md | markdown | 1.8 KB |
references/commands.md
View on GitHub# Command Playbooks (terminal)
Each command takes one named target (a panel, a bar, a card type). Every refine command finishes with `polish`.
## shape (plan before code)
- Ask 2-3 questions per round: purpose, who uses it and how often, what success looks like; then content ranges, states and width limits.
- Output a brief: goal, users and frequency, content inventory with ranges, states, layout at each width, keyboard/mouse model, open questions. No code.
## distill (remove what does not earn its rows)
- One primary goal per panel; say each fact once per screen.
- Flatten nesting; delete group titles that only restate the content.
- Collapse finished or idle content to a summary row.
- Keep the fixed rail; distill its contents, not its existence.
## layout (spacing and rhythm)
- Squint test first; fix order before decoration.
- Group by proximity, then separators, boxes last.
- Tight inside groups, one consistent gap between groups.
- Align values in a column; use tabular alignment for numbers and durations.
- Predictable structure is the affordance: same section shape everywhere.
## typeset (hierarchy without font sizes)
- Define 3-4 roles: title (bold or accent), value (normal), label (dim), meta (dim italic or muted).
- Apply each role identically everywhere it repeats.
- Prose wraps at 60-80 columns; chrome truncates.
## quieter (tone down)
- Fewer background tints, flatter surfaces, fewer accent uses, no decorative motion.
- Keep a few anchors (the accent on the active item) so it still has a point of view.
## bolder (amplify one thing)
- Touch only the named target; amplify what the system already owns.
- Commit to one move and quiet everything around it.
- Skeleton test: the structure must read with placeholder text.
## colorize (roles, not swatches)
- Map every color to a role from the theme; accents only for action, selection and state.
- Add a second neutral for side panels and bars; check contrast on it.
- Pair every state color with a glyph or word.
## clarify (copy)
- Controls name their action; errors say what failed, why, and how to recover.
- Empty states distinguish first use, nothing yet and failure.
- Loading names the real operation; never fake progress.
- One term per concept across the whole shell.
## harden (edge cases)
- Long paths and branch names, emoji and CJK widths, 0/1/1000 items, very long tasks.
- Missing data, failed commands, permission errors: one failing part never blanks the whole surface.
- Resize during render; collapsed and expanded transitions.
## adapt (other sizes)
- Rethink per width, do not scale: wide (rail visible), breakpoint (exactly at and below it), narrow 80x24.
- Decide what moves to the status bar or a key when the rail disappears.
## polish (final pass)
- Classify each drift: missing token, one-off value, concept mismatch, local defect.
- Triage: broken task, missing states, flow drift, visual, cleanup.
- One glyph family; every control has all its states; no stray colors.
- Re-capture and compare with the brief and the previous PNG.
## document (context)
- Write or update `DESIGN.md` per `design-doc.md`; never overwrite an existing one silently.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |