prettymsg
Skill1 starsChanged 4 months agoMITView on GitHub
Repackage the current answer / chat context into a pretty, scannable, bullet-style overview rendered RIGHT HERE in the chat (not a file, not sent anywhere) — TL;DR-first, bold mini-headers, a few tasteful emojis, kept tight. Use when the user says "/prettymsg", "prettymsg <focus>", "pretty msg", "pretty overview", "tldr pretty", "make this pretty (in chat)", "give me a pretty overview", "pretty recap".
In this repository9 other files it gives its agents.
- commitpushSkill
- discordembedSkill
- discordmsgSkill
- endSkill
- handoffSkill
- prettyhtmlSkill
- startSkill
- wrapupSkill
- wtfSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
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
- Very specific: line 39
Use
code / file:linefor anything they'd click or grep. - Distinctive: line 61
If the user actually wants it shared/saved, point them at the right sibling — don't silently do it here.
- Very specific: line 66
Read $ARGUMENTS → pick the subject (the named focus, or the latest substantive result/state in the chat) and honor any tone/length the args …
- Very specific: line 71
Keep it tight; if there's important overflow, *offer* /prettyhtml or /discordmsg rather than inlining a wall.
≈ “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.
Files
/prettymsg — a pretty, skimmable overview, in the chat
Take whatever was just figured out / discussed / produced in this chat and re-render it as a pretty, bullet-style overview the user can read at a glance — output directly in the chat reply. Nothing is written to a file, copied to a clipboard, or posted anywhere. This is the lightest of the overview skills: it's just a beautifully-formatted answer.
Markdown renders in the Claude Code terminal. **bold**, ##/### headers, - bullets, and
plain emoji all display. So "you can't format in chat" is a myth — lean into it.
Who it's for (this sets the tone)
The reader was in this chat — so it's a recap for them, not an explainer for an outsider
(that's /discordmsg). Assume they know the context; the job is to make the current state / answer
instantly scannable: what's the headline, what matters, what's next. No journey, no dead-ends,
no re-teaching.
The format (the rules of the output)
- Lead with a one-line TL;DR. Bold it. One sentence, the headline. They should get the gist from that line alone.
- Then 2–5 bold mini-headers, each with tight bullets under it. Overview first, detail only if it earns its place.
- A few emojis — tasteful, as section anchors / status flags, not confetti. One per header at most, plus the odd inline status flag. Useful set: ✅ done · 🐛 bug · 🔧 fix · ⚠️ caveat · 🚀 shipped · 📊 numbers · 🎯 next · ⛔ blocker · 💸 cost · ❓ unknown. Plain unicode only (exotic ones can render as tofu boxes in some terminals).
- Bullets are short. Fragment > sentence. One idea per bullet. Bold the key term at the front of
a bullet when it helps the eye land (
**Login:** blocked on …). - End with a 🎯 Next line (1–3 bullets) when there's a decision or action pending — that's usually what the reader is scanning for.
- Use
`code`/file:linefor anything they'd click or grep. Keep real numbers exact.
Keep it TIGHT — that's the win
Length is the cost, not emojis (a plain emoji is negligible; use them freely within taste). So
the discipline is brevity: a crisp ~8–16 bullet overview, never a wall. If the underlying thing
is big, summarize and offer "want the full detail / /prettyhtml it?" — don't dump it. A pretty
overview that doesn't fit on one screen has failed at being an overview.
Faithful, not inflated
Summarize only what's actually in the chat. Don't invent status, don't upgrade "unvalidated" to "done", don't bury a ⛔ blocker to make it look tidier. Pretty ≠ rosy. If state is messy, the overview says so — clearly and prettily.
How it differs from its siblings (don't duplicate them)
/prettymsg→ in-chat pretty overview. No file, no clipboard, no send. ← this one./discordmsg→ paste-ready Discord message for an outsider, copied to clipboard (+.txt)./discordembed→ richer Discord embed-style writeup (bigger budget, bold headers), copied./prettyhtml→ writes a polished HTML file and opens it in your browser.
If the user actually wants it shared/saved, point them at the right sibling — don't silently do it here.
Execution steps
- Read
$ARGUMENTS→ pick the subject (the named focus, or the latest substantive result/state in the chat) and honor any tone/length the args request ("just next steps", "super short"). - Compose: TL;DR line → bold mini-headers with tight bullets → 🎯 Next. Emojis as anchors,
bullets short, numbers exact,
file:linewhere useful. - Output it directly in the chat reply — no tool calls, no file, no clipboard, no posting.
- Keep it tight; if there's important overflow, offer
/prettyhtmlor/discordmsgrather than inlining a wall.
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
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.

