skillZs
★ LIVE SKILL TAGS ★
>>> LIVE SKILLS INDEX <<<
* OPEN SOURCE *
NO LOGIN, NO TRACKING
※ REAL INSTALL DATA ※
← back to all skills
saadshahd/moo.md72 installs

explain

Show one settled thing so the user holds it. Use when the user does not hold something already settled.

How do I install this agent skill?

npx skills add https://github.com/saadshahd/moo.md --skill explain
view source ↗

Is this agent skill safe to install?

  • Gen Agent Trust Hubpass

    The skill is generally safe but includes instructions that involve processing untrusted data from git commits and pull request threads, which creates a surface for indirect prompt injection.

  • Socketpass

    No alerts

  • Snykwarn

    Risk: MEDIUM · 1 issue

What does this agent skill do?

Explain only what the user has said or shown they don't hold, and only from what you have read: retrieve it first, explain what it holds, and name what it leaves out as a gap. Nothing in reach holds it → name what's missing, ask for it, and wait. Why something is this way is also read from the commits and PR threads that brought it in and changed it: quote the reason given there, and mark your own inference as one.

Open with the thing's common name, where it has one, and what it is in general, then what it is here. Answer in the fewest small views that make the key point, each under its own one-line caption, in the shape the parts already have: code that runs as a call tree of its function names, logic as pseudocode, UI as a component tree, files as a file tree, a change as a diff on that view, a spatial idea — layout, overlap, scroll position, a before and after — as an SVG picture opened for the user, anything else as a tree, numbered steps or a table. Three or more moving parts → a series of views, each redrawing the last with one part added. Put each text view but a table in a code fence tagged with its language — diff, tsx, text — so the terminal colors it. A line holds a name from the source, with a # few words comment at most. Mark undecided parts [open]. Then wait.

Each pull — a question, a ruling, an "I still don't get it" — is answered the same way, on the pulled part alone. "I still don't get it" means the view was wrong: change the view, never add words.

It ends when the user says it back in their own words or acts on it — claim it landed only then — or moves on, where the answer stands as given.

The user holds it and you're the one missing it → ask a direct question. The user holds it but cannot yet put it into words → use elicit skill. The user wants a verdict on it, not an understanding of it → use judge skill.

Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.

<a href="https://skillzs.dev/skills/saadshahd/moo.md/explain">View explain on skillZs</a>