wong-setup
Skill7 starsChanged 4 days agoMITView on GitHub
Evaluate or install WongStack from any folder, Cloudflare hosting included, from one token. Installed repos use /wong-sync.
In this repository14 other files it gives its agents.
- AGENTS.mdAGENTS.md
- accessSkill
- browserSkill
- dailySkill
- memorySkill
- agent-browserSkill
- connectSkill
- improve-codeSkill
- lessonSkill
- paste-this-into-your-aiSkill
- reply-linkSkill
- uiSkill
- update-dependenciesSkill
- wong-syncSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
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
- Very specific: line 25
Before cloning/writing, read the selected source's setup/tools under that raw root.
- Distinctive: line 29
Use a separate clean source checkout, reading that reference there.
- Distinctive: line 33
Before writing, ask whether they have the Cloudflare user token, explaining its purpose and selected source's link/fallback.
- Distinctive: line 41
Run Step 1 of the source's provisioning runbook.
≈ “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
/wong-setup
Pick the folder
Select before writing; preserve existing projects:
- The open folder has
.claude/.wong-stack.json→/wong-sync. - It is empty, or holds only a
.gitwith no commits → install here. - Otherwise try
~/wongstack, then numbered siblings. Use a missing/empty folder; an install record routes to sync.
Name the target by its absolute path. Create it after tools/token.
Get the computer ready
Select the requested GitHub repo, else https://github.com/matthewwong525/WongStack. Normalize its HTTPS owner/repo URL. Read its default branch through GitHub's API, authenticated if needed. Use https://raw.githubusercontent.com/<owner>/<repo>/refs/heads/<default branch>; never assume main.
Route: on Mac or Linux, unless GitHub is asked for, follow the Artifacts route. Both need Cloudflare's paid plan.
Before cloning/writing, read the selected source's setup/tools under that raw root. Follow its links, .nvmrc and prerequisites. Missing source/requirements stop; never substitute another stack.
Get the latest source
Use a separate clean source checkout, reading that reference there. Install its version/commit into the target. When installing, start learning from past chats now.
Get the Cloudflare token
Before writing, ask whether they have the Cloudflare user token, explaining its purpose and selected source's link/fallback. Without it, wait. Request no value until Git protects its file.
Install through the normal workflow
<a id="step-2--deep-research-the-target-repo"></a>
Work in the target using source-checkout skills/references, starting with explore. Setup stays source-only.
Run Step 1 of the source's provisioning runbook. Commit and push only after installation.
Invoke /explore with this description, filled in, plus the user's intent:
Set up WongStack in
<target>from<selected repository>at<source path>, version<version>, commit<commit>. Use a reachable owner login email separately from git authorship. Private pages and previews are automatic when the account allows them; ask no enable-or-public question. An account needing a card finishes open and recommends the card as optional. Put the files in place as one task: node<source path>/.claude/skills/wong-setup/scripts/install.mjs --source<source path>--target<target>--repo<selected repository>; write no copy step of your own. Ifpaseois installed, runnode .claude/skills/daily/skills/schedule/scripts/presets.mjs addin<target>; if they took the helper,node .claude/skills/reply-link/scripts/helper.mjs install; no failure stops setup. Run Steps 2–5 of<source path>/.agents/skills/wong-setup/references/cloudflare.md. Continue through the requested stage.
Explore/plan/apply/save own questions/planning/installation/first deploy. Install every component without asking; use shared asks.
<a id="step-7--bootstrap-seed-hand-off"></a>
install.mjs copies every payload category, makes the agent folder and its links, the wiki hubs, ignore rules, install record and openspec/, and prints one JSON report. A stop exits 1 with its cause and leaves the folder as it found it; fix it by the failure map, then run it again.
Evaluation stops in exploration. Installation continues through plan/apply/save without another interview or approval. After first deploy, give Step 5's report, naming preset additions/keeps/skips or unavailable Paseo.
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.

