explain-clearly
Skill6 starsChanged 6 days agoMITView on GitHub
Explain a concept clearly in the shape the subject has. Use whenever the user asks 「X 是什么」「为什么需要 X」「A 和 B 能不能/怎么通信」「给我讲讲 X」「用大白话解释 X」「帮我理解 X」「容易看懂」, \"what is X\", \"how does X work\", \"why does X exist\", \"explain X\" about mechanisms, architecture, protocols, or processes — even if they never say \"explain\". Do NOT use for hands-on tutorials (teach), marketing articles (article-writer), or design-doc rewriting (tech-review-doc).
1 other project uses a version of this.
Adopted by
Most recently taken up by
- BananaSoldier01/explain-clearlytheir own version, first seen Oct 13, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- BananaSoldier01/explain-clearly★ 3same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository79 other files it gives its agents.
- AGENTS.mdAGENTS.md
- openspec-apply-changeSkill
- openspec-archive-changeSkill
- openspec-bulk-archive-changeSkill
- openspec-continue-changeSkill
- openspec-exploreSkill
- openspec-ff-changeSkill
- openspec-new-changeSkill
- openspec-onboardSkill
- openspec-sync-specsSkill
- openspec-verify-changeSkill
- ai-proxy-disciplineSkill
- analysis-coreSkill
- android-webview-debugSkill
- article-writerSkill
- astrill-controlSkill
- brainstorm-workflowSkill
- browser-debug-toolkitSkill
- clarifying-question-disciplineSkill
- code-design-reviewSkill
- competitive-researchSkill
- completion-evidence-disciplineSkill
- context-budget-disciplineSkill
- decision-fog-disciplineSkill
- delivery-disciplineSkill
- design-approval-gateSkill
- domain-language-disciplineSkill
- effective-web-researchSkill
- essence-diagnosisSkill
- evolution-reviewSkill
- feature-branch-closeoutSkill
- figma-pixel-implementSkill
- figma-pixel-verifySkill
- git-branch-protectSkill
- git-commitSkill
- git-conflict-resolveSkill
- git-release-finishSkill
- git-release-startSkill
- git-worktree-disciplineSkill
- goal-driven-queueSkill
- goal-driven-workflowSkill
- go-deploySkill
- hybrid-debugSkill
- intake-interview-disciplineSkill
- jira-fix-queueSkill
- jira-fix-workflowSkill
- jira-readSkill
- jira-status-lifecycleSkill
- jira-status-writebackSkill
- jira-wiki-markupSkill
- known-issue-researchSkill
- learn-and-improveSkill
- merge-disciplineSkill
- multi-agent-debateSkill
- node-version-disciplineSkill
- opsx-jira-fix-queueSkill
- opsx-jira-fix-workflowSkill
- opsx-solve-workflowSkill
- opsx-workspace-gateSkill
- perf-optimize-workflowSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Use this for concept and mechanism questions: what something is, why it exists, and whether two parts can communicate.
- Distinctive: line 24
When the user has already named a form, use that form.
- Distinctive: line 48
No such shape: leave the visual out.
- Very specific: line 54
Create the workspace-root directory .explain-clearly/ when a file form needs it.
- Distinctive: line 56
Do not edit the workspace gitignore.
≈ “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
Explain Clearly
Explain a technical concept so the user can see how it works. Give the relevant causes in the shape the subject has. An analogy is optional and must not invert a constraint. Deliver the explanation in the user's language, in the form they chose.
When to use
Use this for concept and mechanism questions: what something is, why it exists, and whether two parts can communicate. IPC, memory isolation, event loops, virtual DOM, consensus, and caching are typical subjects.
Route elsewhere and stop when the request is:
- A hands-on tutorial or onboarding ("教我怎么用 X", "walk me through setting up X") →
teach - Rewriting existing docs into an article →
article-writer - Turning a design doc into a review doc →
tech-review-doc
Choose the form
When the user has already named a form, use that form. When they have not, ask once. Offer exactly these choices, and recommend HTML:
- current session
- HTML
- Markdown
- plain text
What every explanation contains
Name the subject in one sentence under the title. The shape under Visuals is the whole explanation.
Include a fact only when the picture cannot show it, when it is evidence for that cause, or when the user is choosing an action.
- Evidence — cite files, line numbers, and APIs read in this session, next to the step or part they prove. If the concept has no footprint in the current codebase, use a canonical public example and label it as such. If no canonical public example exists, say that no concrete example is available. An invented path, line, or API is a failed explanation.
- A choice — state the design implication only when the user is choosing an action.
- An analogy — one scene, only when it keeps every defining constraint. Omit the analogy when the picture already shows that relation. If a constraint would be lost or inverted, adjust or replace the analogy before delivery.
Visuals
Apply the first match. Draw only relationships that exist.
- A role-handoff flow: the diagram is the explanation and the diagram leads the page. Distinguish the roles, show order, and put the handed-off items on the diagram. A comparison table does not satisfy that flow. Do not also require the step player. Put the sentence the picture cannot show next to the part of the diagram it belongs to.
- The subject happens step by step in time: lead with those steps. HTML shows one step at a time, opens on the first step, provides previous and next, and does not autoplay. HTML may contain one inline script, and that script only operates the player. Markdown, plain text, and the current session show the same steps as a numbered sequence and do not add those controls. A position mark is not a message. Each step shows only what is true at that moment. Put the sentence the picture cannot show next to that step. Put the code that proves a step on that step.
- Comparison or mapping, and not time-ordered: one table leads, one row per item. Topology or flow, and not time-ordered: one static diagram leads. Put the sentence the picture cannot show next to that table or diagram.
- No such shape: leave the visual out. A few sentences of the relevant cause are the explanation.
HTML and Markdown include that visual when the concept has a shape. Plain text stays prose. The current session uses the same first match: a diagram for a role-handoff flow, a numbered sequence only when the subject is time-ordered and not a role-handoff flow, a table or diagram for the other shapes, and no visual when the concept has no shape.
Deliver
Create the workspace-root directory .explain-clearly/ when a file form needs it. Name each file <topic-slug>-<local-YYYYMMDD-HHmm> with the extension for that form. Write a new file every time. Leave an existing file in place.
Do not edit the workspace gitignore.
- current session — put the explanation in the chat. Include the visual when the concept has a shape. Write no file.
- HTML — write
.explain-clearly/<name>.htmlin the user's language. Readassets/explanation-shell.htmlas a reference for a page: inline CSS, and text written as text. The page is a title, one sentence under the title, and the visual for that shape. When the caller fixes the form to HTML and supplies an evidence-page path, keep the title, the one sentence, and the visual, and add one link after the visual. The link text is 证据留痕. The href is the relative path from this file to the evidence page. The link does not replace the visual. When the caller does not supply an evidence-page path, do not add that link. The shell itself has no script. The written page has no script except one inline script for a time-ordered player. The English words in that file are examples, because this skill is written in English. Omit the visual block when the concept has no shape. Open that local HTML file in the browser. In the chat, give the file path and ask the user to view the page. Do not paste the full explanation. If the browser cannot be opened, still give the file path. - Markdown — write
.explain-clearly/<name>.mdin the same shape as the visuals rules. Include the visual when the concept has a shape. In the chat, give the file path and do not paste the full explanation. - plain text — write
.explain-clearly/<name>.txtas prose only. In the chat, give the file path and do not paste the full explanation.
Worked examples
Two calibrated examples live in reference.md. Read them to match depth and to see how a generic example is labeled when the codebase has no footprint.
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.

