agentleFS
solo-istsolo-ist/prose/.claude/skills/screenshots/SKILL.md

screenshots

Skill8 starsChanged 5 months agoMITView on GitHub

Generate the README hero + Mac App Store marketing screenshots (retina, Prose theme, light/dark) via Playwright.

  • Reads credentials

18 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

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

  • 0

    Common

    Most of the 18 edited versions we read keep these.

  • 0

    Distinctive

    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:

    Don't try to wire a live call.

    line 89
  2. Very specific:

    If a scene is added/removed/renamed there, update the spec scenes and this table, regenerate, and replace the files in …

    line 103

“Common” is measured: kept by at least 6 in 10 of the 18 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

Marketing Screenshot Generator

Produces the README hero and the Mac App Store screenshot set from a clean, isolated, seeded Prose instance — deterministic, retina (2560×1600), no private data, no live API calls. Built on the e2e Playwright harness.

What it produces

OutputScene (App Store copy)Mode
01-hero-editor.png (also the README hero)Hero — post + Projects explorer☀️ Light
02-ai-chat.pngAI Alongside You — chat + mode-switch card🌙 Dark
03-review-diffs.pngReview Every Suggestion — Quick Review☀️ Light
04-edits-history.pngEdits History — the Activity panel🌙 Dark
05-projects-favorites.pngProjects & Favorites — Favorites view☀️ Light

Light hero, then alternating dark/light. Scenes mirror the shot list in docs/app-store-copy.md. Keep that copy and this generator in sync when the shot list changes.

Files

FilePurpose
e2e/electron.screenshots.spec.tsScenes 1, 3, 4, 5 — isolated profile, seeded projects/favorites + the blog fixture
e2e/electron.screenshot-chat.spec.tsScene 2 — the chat mode-switch, injected deterministically (no live AI)
e2e/fixtures/a-chorus-of-human-voices.mdDemo document (a real blog post). Frontmatter is stripped at runtime so the title leads; a few obvious typos are injected below the fold for the review/activity demos
docs/app-store-screenshots/Destination for the App Store set
docs/images/prose-screenshot.pngDestination for the README hero (= scene 1)

Both specs are excluded from the normal e2e run — they self-skip unless SHOOT=1 is set (test.skip(process.env.SHOOT !== '1', …)).

Run

# 1. Free port 9222 — the dev build binds it; a running dev server blocks the
#    screenshot instance. Stop it first (PID file), restart when done.
kill "$(cat .dev.pid)" 2>/dev/null

# 2. Build (the specs launch the built app at out/main/index.js).
npm run build

# 3. Generate → writes to /tmp/prose-shots/
SHOOT=1 npx playwright test electron.screenshots electron.screenshot-chat --workers=1

# 4. Review every shot before committing (they're marketing assets).
open /tmp/prose-shots/*.png

# 5. Place into the repo.
cp /tmp/prose-shots/0{1,2,3,4,5}-*.png docs/app-store-screenshots/
cp /tmp/prose-shots/01-hero-editor.png docs/images/prose-screenshot.png

Inspect each PNG (e.g. with the Read tool) and iterate — re-run a single scene with ... electron.screenshots -g "03 review" --workers=1.

How it works

  • Isolated profile — each spec seeds a throwaway PROSE_USER_DATA_DIR with settings.json (Prose theme color: 'prose', projects/favorites, a fake llm.apiKey so the chat renders configured, aiConsent). No private data, reproducible.
  • Retina — the window is set to 1280×800 logical; on a retina Mac that captures at 2560×1600 (DSF 2), the App Store spec. Verify with sips -g pixelWidth -g pixelHeight.
  • Theme — color: 'prose' (the gold theme; not the mono default). Light/dark per scene via the dark class on <html>.
  • Seeded content via test seams (always-on, see src/renderer/main.tsx):
    • window.__prose_tools.executeTool(name, args, mode, provenance?) — drive open_file, suggest_edit (pass provenance.model so the Activity panel shows a real model, not "Unknown"), add_comment, accept_diff.
    • window.__prose_chat — seed a conversation + inject messages. Scene 2's mode-switch is an assistant message whose content carries a real <tool-result name="request_mode_switch" success="true">{json}</tool-result> tag, so the rendered card is the actual RequestModeSwitchResult component.

Gotchas

  • Live AI can't be automated. The real API key lives in the OS Keychain; safeStorage is unavailable to the Playwright-launched Electron, so a fresh automated run can't read it. Scene 2 is therefore mocked (the card is the real component with scripted content) — visually identical to a live exchange. Don't try to wire a live call; if you must, you'd seed a plaintext key into the profile's settings.json (ipc.ts reads settings.llm.apiKey first).
  • Port 9222. The unpackaged build runs in dev mode and binds the DevTools port; a running dev server (or a second screenshot run) collides. One at a time.
  • Don't open -a Prose while generating — it launches the installed app, which shares userData on the case-insensitive filesystem and fights the single-instance lock.
  • Demo content — to change the document, replace the fixture (keep it a real, on-brand piece). Typos for the review/activity scenes are injected at runtime in electron.screenshots.spec.ts (heroBody.replace(...)), not in the fixture.

Updating the App Store set

The shot list is owned by docs/app-store-copy.md (issue #612 owns the assets). If a scene is added/removed/renamed there, update the spec scenes and this table, regenerate, and replace the files in docs/app-store-screenshots/.

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.