agentleFS
sefaertuncsefaertunc/anthropic-watch/.claude/skills/frontend-design-system/SKILL.md

frontend-design-system

Skill1 starsChanged 7 months agoMITView on GitHub

Dashboard conventions — single static HTML file, vanilla DOM APIs, strict XSS hygiene

9 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 3 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

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

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

    URLs must be validated before use: Use the safeUrl() helper (rejects anything that isn't http:// or https://) before setting href on any …

    line 18
  2. Very specific:

    If a feature needs per-source items, fetch feeds/{source-key}.json separately.

    line 31
  3. Distinctive:

    Keep styles surgical: don't restyle unrelated sections when editing.

    line 39
  4. Distinctive:

    No explicit WCAG level targeted — it's a developer tool.

    line 43
  5. Very specific:

    Run the pipeline locally (npm start) to populate public/feeds/.

    line 53
  6. Distinctive:

    If browser tests ever become necessary, introducing them is a separate project decision — don't add a test framework as part of an …

    line 58

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “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

Frontend Design System — anthropic-watch

The "frontend" is one file: public/index.html. It is a static dashboard that fetches run-report.json and all.json from GitHub Pages and renders them with vanilla DOM APIs.

There is no framework (no React/Vue/Svelte), no build step (no bundler, no transpiler, no PostCSS), and no component library. These are intentional choices — the dashboard must be viewable by cloning the repo and opening the file, and deployable by copying the file to Pages.

Rendering Rules

  1. Never use innerHTML for anything that includes data. Build the DOM with document.createElement(...) and assign text via textContent. The v1.0.1 release replaced all innerHTML string templates specifically because they allowed javascript: URL XSS through item.url.
  2. URLs must be validated before use. Use the safeUrl() helper (rejects anything that isn't http:// or https://) before setting href on any element. This applies to item.url, sourceName links, anything that comes from the feed.
  3. External item links use rel="noopener noreferrer" and target="_blank".
  4. textContent over innerText for consistency and performance.
  5. Dates: format with toLocaleString() or equivalent at render time; never trust pre-formatted date strings from the feed (they are ISO 8601, not display-ready).
  6. Null handling: item.date may be null; render a sensible placeholder (blank or dash), never "null" or "undefined".

Data Sources

The dashboard fetches two files from the same origin (GitHub Pages):

  • feeds/run-report.json — for source status, health, timing
  • feeds/all.json — for the recent-items list

If a feature needs per-source items, fetch feeds/{source-key}.json separately. Never combine: the schemas are intentionally distinct (see docs/FEED-SCHEMA.md).

Handle fetch failures gracefully — GitHub Pages can have brief outages. Showing a stale-but-cached view is better than a blank page.

Styling

Inline <style> block at the top of public/index.html. No external stylesheets, no Tailwind, no CSS frameworks. Design tokens (colors, spacing) are defined as CSS custom properties in :root if present, or inlined where used.

Keep styles surgical: don't restyle unrelated sections when editing. The dashboard is utilitarian by design — dense info, no animations, no marketing polish.

Accessibility

No explicit WCAG level targeted — it's a developer tool. Keep basics sane:

  • Semantic HTML (<table> for tabular data, <button> for actions, <a> for links)
  • Visible focus states
  • Color contrast sufficient for status dots (green/amber/red must be distinguishable by more than color alone — use shape, label, or hatching)

Testing

No automated browser tests. Verification is manual:

  1. Run the pipeline locally (npm start) to populate public/feeds/.
  2. Open public/index.html in a browser.
  3. Visually confirm the changed surface renders correctly.
  4. For security-sensitive edits (anything touching URL handling, DOM rendering, or user-controlled content), add an explicit note to the PR describing what was checked.

If browser tests ever become necessary, introducing them is a separate project decision — don't add a test framework as part of an unrelated change.

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.