agentleFS
Tamircohen28Tamircohen28/tamirs-superpowers/skills/documentation/dark-terminal-doc/SKILL.md

dark-terminal-doc

Skill2 starsChanged 3 months agoMITView on GitHub

Use when a single-file HTML technical document should carry a dark developer / terminal aesthetic — monospace, terminal-green on near-black. Trigger on: "dark terminal doc", "terminal-style HTML", "make it look like a terminal", or a comparison table, reference sheet, changelog, release notes or API cheatsheet in that look, or one matching a document already in this style. Not for a doc with no stated aesthetic.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Always produce a single self-contained .html file: No external CSS files, no JavaScript frameworks, no CDN scripts beyond the Google …

    line 84
  2. Very specific:

    Never use solid backgrounds for tags or badges: All fills must use rgba with 12–15% opacity so the dark --bg bleeds through.

    line 85
  3. Distinctive:

    Max-width is 1200px throughout,: not 960px — comparison tables need horizontal breathing room.

    line 92
  4. Distinctive:

    Before handing off the HTML file, verify:

    line 115

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Dark Terminal Document Design System

Why this skill exists

Standard markdown renders poorly in meetings, portfolios, and async handoffs. Notion exports are generic, Confluence is ugly, and spinning up a React/Next.js site for a one-off comparison table is overkill. This skill produces a single self-contained HTML file — no dependencies, no build step, no framework — that looks like a polished internal developer tool: dark IDE aesthetic, dense tables, semantic status colors. The output opens directly in any browser, can be emailed as an attachment, or dropped into GitHub Pages.

Naive approaches fail because: ad-hoc inline styles drift from the design system, font loading order breaks the monospace/sans split, and brand accent colors used directly in body text create contrast issues (they need pastelizing).


Workflow

  1. Confirm the document type from the user's request:
    • Comparison doc (two subjects side by side) → use all columns + summary cards
    • Reference/cheatsheet (one subject, many rows) → Command / Syntax / Description
    • Changelog/release notes (chronological) → Change / Impact / Description
    • Feature breakdown (one product, feature × status) → Feature / Status / Notes
  2. Identify brand colors for comparison docs. For single-subject docs, pick one --brand-a accent and omit --brand-b.
  3. Load the CSS design system from references/css-design-system.md and copy the token block verbatim. Override only --brand-a and --brand-b.
  4. Use the HTML skeleton from references/html-skeleton.md. Fill section headers with emoji prefixes. Every row that has a clear winner gets a winner-badge after the explanatory text in the notes column.
  5. Write the file as a single .html file using the Write tool. No external CSS files, no JS frameworks, no CDN scripts (Google Fonts import is the only external call — degrades gracefully offline).
  6. Verify by describing the output structure to the user — number of sections, rows, and which brand wins the summary (for comparison docs).

Hard Rules

  1. Always produce a single self-contained .html file. No external CSS files, no JavaScript frameworks, no CDN scripts beyond the Google Fonts import.
  2. Never use solid backgrounds for tags or badges. All fills must use rgba with 12–15% opacity so the dark --bg bleeds through.
  3. Brand accent colors (--brand-a, --brand-b) must be pastelized for body text in data cells. Using the full-saturation color directly on body text causes glare and contrast issues.
  4. Winner badges belong in the notes column, after the explanatory text — badge is a visual summary, not a lead-in.
  5. Never use light backgrounds, Inter/Roboto fonts, purple gradients, glassmorphism, or rounded hero cards. This is a terminal aesthetic, not a marketing page.
  6. Never drop the generic font fallback. 'Syne', sans-serif and 'IBM Plex Mono', monospace keep the page readable when the Google Fonts import is blocked — offline, behind a proxy, or under a strict Content-Security-Policy. The import is the one permitted external call precisely because it degrades; a bare font-family: 'Syne' does not.
  7. No harness features. The output is a .html file opened from file://. It requires no agent runtime, no skills, no subagents, no build step — and must keep requiring none.
  8. Summary cards are mandatory for comparison docs — three boxes: brand-a wins / brand-b wins / tie. Single-subject docs (cheatsheets, changelogs) may omit them.
  9. Max-width is 1200px throughout, not 960px — comparison tables need horizontal breathing room.
  10. Section headers must use text-transform: uppercase, letter-spacing: 2px, and an emoji prefix to create visual rhythm and scanability in long tables.
  11. Diagrams use the canonical dark Mermaid palette — the init line from config/docs/mermaid-init.txt per rules/dev/documentation-standards.md — unless the user asks for another palette for this document.

What NOT To Do

WrongRight
color: #D97757 in a table data cellcolor: #EBA98C (pastelized)
.tag { background: #4CAF7D; }.tag-yes { background: rgba(76,175,125,0.15); }
<span class="winner-badge">wins</span> Context is betterContext is better <span class="winner-badge winner-brand-a">brand-a wins</span>
Splitting CSS into a separate .css fileAll styles inline in <style> block
font-family: Inter, sans-seriffont-family: 'Syne', sans-serif
box-shadow on .table-wrapborder: 1px solid var(--border) only
border-radius: 8px on .table-wrapborder-radius: 16px
Prose descriptions using IBM Plex MonoProse in Syne; metadata/labels in IBM Plex Mono

Quick-Reference Checklist

Before handing off the HTML file, verify:

  • Single .html file — no external dependencies except Google Fonts import
  • CSS variables match the token reference exactly (--bg, --surface, --surface2, --border, --text, --muted, --yes, --no, --partial, --brand-a, --brand-b)
  • --brand-a and --brand-b overridden to match the document's subjects
  • Table header columns use th-brand-a, th-brand-b, th-meta classes
  • Data cells use col-brand-a/col-brand-b with pastelized hex values, not full-saturation vars
  • Section headers: uppercase, 2px letter-spacing, emoji prefix
  • Tags use rgba fills (not solid colors)
  • Winner badges placed after explanatory text in notes column
  • Summary cards present (for comparison docs only)
  • Subtitle under h1 in IBM Plex Mono with // title · date pattern
  • max-width: 1200px on all top-level containers
  • border-radius: 16px on .table-wrap

Reference Files

FileWhen to load
references/css-design-system.mdEvery document — copy the full CSS token block and component library
references/html-skeleton.mdStarting point for every document — choose column layout by doc type

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.