LSS Everything: AI Skills for Busy Business Owners

LSS Everything is a pack of ready-made steps that help an AI assistant do work for your business. Use it to turn a video into posts, review your results, or check your website. It is part of our business growth system; start with one job and a draft you can check.

Choose one job → Use its steps → Check the draft → Save and reuseFollow the numbered steps in order. The path runs across the first row, returns to the lower left, then continues right.Choose onejobUse itsstepsCheck thedraftSave andreuse
Start with one job. Check the result before you use it or repeat it.

Choose a business task · What v19 means · Install and test LSS Everything

What can LSS Everything help your business do?

Local Service Spotlight (LSS) built this pack for owners and small teams who want a clear way to get repeat work done. A skill is a set of steps for one kind of task. You give the AI the facts, choose the task, and check what it makes. You do not need to learn all the skills at once.

Your jobWhat to ask forWhat to check
Reuse a real videoAn article draft and short posts based on your recording.The words match the source and give your audience a useful next step.
Review the week’s resultsA report of your numbers, what they mean, and what to do next.The numbers match the records you provided. Missing facts stay marked unknown.
Find website problemsA list of issues that may make your site harder to find or use.Each issue names the page, the evidence, and a useful fix.
Review software costsA list of tools you pay for, with possible overlaps to review.The costs match your records. You decide what to keep or cancel.

These are example outputs, not promises of more sales or lower costs. The pack supplies the methods. Your sources, the tools you allow, and the checks on the result make those methods useful.

Start with one useful result

Have ready: your business website, one task, the facts it needs, and a person who will check the draft. Start a work record when you begin and add the result as you go. Set a Goals, Content, Targeting brief: the result you want, the material you have, and who it is for.

  1. Choose one job. If you are unsure, use the first-result prompt to review your public business page and draft one improvement. That first exercise uses a public guide; it does not require installing the pack.
  2. Set up the method when needed. Follow the install guide for your app. Check that the chosen skill is installed, enabled, and actually loaded in a fresh task. Add only the files or account connections that job needs.
  3. Check the result. Open the draft, check its facts against the sources, and confirm it answers your request. Save what passed and what still needs work.
  4. Choose what happens next. Have the owner approve the next use. For repeat work, agree on the schedule and where results go, then check the first scheduled run. A one-time task needs no schedule.

You are done with the first test when you have one useful draft, checked sources, and a clear next action. Save a work record, or meta article, so the next run can learn from this one.

What does “LSS Everything v19” mean?

In our September 22, 2026 account check, Claude showed Lss everything as enabled, labeled v19, with 36 skills. That is the version label we observed in Claude. The marketplace uses a separate version number: its metadata says 1.6.0 and lists 36 skills in the source checked on September 23. These two labels do not establish a one-to-one release match.

You do not need to memorize either number. Open your installed plugin, check the listed skills, then test the one you need in a fresh task. Use the agent update guide when your copy needs refreshing. A version label alone does not show that a job ran.

The article process for any agent, ours or yours, lives at Article guidelines. You do not need the whole plugin to follow it. The plugin is one way to load the process into your app; updates still need a fresh check.

What a plugin is, in one paragraph

A plugin is a box. Inside the box are skills. A skill is one job written down to a standard — “audit this site’s SEO,” “harvest my positive mentions” — written so an AI worker can follow the same method each time. Some jobs need your review or approval. The installed version provides its listed skills. Later releases need an update in your app and a fresh activation check.

Ask in plain language. If the skill is not selected, use the app’s skill picker and check that the intended skill loaded. A reply alone does not prove activation.

SkillSteps for one kind of task. A skill does not run itself.
Skill packA group of skills chosen for a type of work. A guide ZIP does not install them.
AccessPermission to use the files and tools this job needs. Record what was granted and tested.
JobOne assigned run with inputs, checks, and a result.
PluginThe box the skills ship in. You install this.
AgentThe worker who runs a skill. You hand it a job, not a script.
Scheduled jobA saved job with a time or supported event. It still needs access and a checked result.
ReceiptA dated record of what ran, its result, checks, and any failure.

Owning the cookbook is not dinner

These seven checks describe different evidence. Installation applies to packaged skills; a schedule is optional for work that needs to repeat. Each claimed state needs its own proof. A green light on step 2 tells you nothing about step 6.

1
AvailableThe skill exists in the pack. Choose the one your task needs.
2
InstalledThe named plugin is installed in the actual account. Check its enabled state and selected skill separately.
3
ConnectedThe agent can reach what it needs — Search Console, the CMS, GA4, the ad account.
4
Scheduled, if neededAn approved job exists with a time or supported event, and a place the result goes.
5
RanIt fired. There is a timestamp and either an output or an honest error.
6
VerifiedSomeone opened the output and it was correct. Not “the agent said it went well.”
7
OutcomeA call, a booking, a sale — traced back to the thing the agent made.
Say the number. “We are at 2 on all 22 agents” is a useful sentence. “The agents are live” is not — it is true at state 1 and at state 7, so it means nothing in between.

What is inside

The marketplace source checked on September 23, 2026 lists 36 skills in lss-everything. Install lss-everything and you get all of them — that is what most people should do, because the skills are written to hand work to each other. The four smaller bundles exist for people who only want one area.

BundleWhat it covers
lss-everythingAll 36 in the reviewed source. Start here.
authority-and-reputation8 — audit intake, Knowledge Panel, AI search, reviews, proof
content-engine7 — articles, video, repurposing, distribution
client-operations13 — onboarding, goals and sources, account access, reporting, audits
quality-and-standards8 — verification, quality checks, judgment, closeout, system principles, and the registry

The Skill Pack Library lists workshop and specialty downloads with their build dates. For this plugin’s included skills, check the reviewed marketplace list. A dated ZIP is a separate snapshot; downloading it does not install or run a skill.

How it updates

The plugin comes from one address: https://github.com/dennisyu/local-service-spotlight-skills. When we improve a skill we change it there. You refresh from the same address — follow that app’s update path and verify the new release. A ZIP of guides is a separate dated download.

Be precise about this one. Whether that refresh happens on its own or needs a click depends on your Claude surface and settings. Do not assume a release reached you. Refresh, then run one real request in a fresh chat and read what comes back. That is the only proof that counts.

One GitHub repository can expose several plugin cards because the marketplace is a catalog. In our pack, lss-everything selects all 36 skills in the reviewed source; the four topical bundles are smaller, sometimes overlapping selections, so choose the all-in-one bundle or selected topical bundles. The five-card and GitHub permission guide shows the real screen, explains plus versus gear, and gives the least-privilege setup.

The task method can travel between AI apps; the package format, supported tools, and access need to be checked in each one. Claude is the install described on this page. ChatGPT and Grok need their own setup and fresh run evidence. The Grok desk is How I use Grok Bot. For the cross-model view, see how the same skill library maps across Claude, ChatGPT/Codex, and Grok.

What a link cannot do

A link to this page does not install a plugin in your account. Add the reviewed package through the app’s supported controls with the required account authority. If authentication or a new grant is needed, record that exact missing step. A copied URL is not proof that setup is complete.

On Claude Team or Enterprise, an owner can use the supported organization plugin controls. A plugin ZIP upload and a GitHub-synced marketplace are different paths; the documented organization sync requires a private or internal repository. Confirm installation, enabled skills, required connections, and one fresh result for each account. Owner provisioning does not prove all of those states. See the official organization setup guide.

What you need

ProClaude documents plugins and Cowork scheduled tasks for paid plans. Confirm current features and workspace controls in your account.
MaxPlans have different usage limits. Test your workload; a higher plan does not guarantee completion.
Team / EnterpriseSupports owner-managed distribution, subject to workspace controls. Confirm the supported package path and each account’s activation and access.

Where the pack fits in the Content Factory

The Content Factory is our process for turning real work into useful content. Choose the skills for the stage you are working on; the pack itself is not another stage.

1. Produce →Record real work and gather proof.
2. Process →Turn sources into drafts and clips.
3. Post →Publish approved work and connect related pages.
4. PromoteShare useful work and measure the result.
Follow the numbered stages. Use what the result teaches you to improve the next task.

The Task Library holds the task recipes and their checks. LSS Everything packages a selection of reusable methods. A work record shows what happened when you used one. To plan the wider setup, follow the AI agent implementation guide.

Next

Start with one result → Browse workshop and specialty packs →

Canonical page. If you arrived from dennisyu.com, blitzmetrics.com or one of the Spotlight sites, this is the source they point at — so there is one description of the system rather than five that drift apart. Package inventory checked against marketplace metadata 1.6.0 at source commit 7d85ec7b, September 23, 2026. The September 22 Claude v19 observation is separate account evidence. Actual task activation and results still need their own checks. Platform claims checked against current official documentation: Use plugins in Claude, Manage plugins for your organization, Schedule recurring tasks in Cowork.

Scroll to Top