Get Found Online
Skill8 starsChanged 5 months agoMITView on GitHub
A bounded "are you findable, and what to fix first" audit for a local or service business, ending in a prioritized fix list you can action this week. Reads your key pages and local search presence, orders fixes by what moves the needle, and hands you the exact change to make. Keyless; rank tracking is the connected-tier upgrade.
In this repository101 other files it gives its agents.
- templates/CLAUDE.mdCLAUDE.md
- Add a FieldSkill
- Assemble Content PackSkill
- Assemble PackSkill
- Audit My AutomationsSkill
- Audit My DataSkill
- automate-thisSkill
- Brand My WorkspaceSkill
- Break Down A ChannelSkill
- Build Brand StrategySkill
- Build Customer VoiceSkill
- Build DocumentSkill
- Build FormSkill
- Build Knowledge Base From DocsSkill
- Build My ProofSkill
- Build My VoiceSkill
- Build Social StrategySkill
- Build SpreadsheetSkill
- Build Work Order ProcessSkill
- Cash Flow ForecastSkill
- Coach My CallsSkill
- Compare DocumentsSkill
- Connect A ToolSkill
- Delegate This WorkSkill
- Describe A ProductSkill
- Design My ScenesSkill
- Design Nurture SequenceSkill
- draft-replySkill
- Email Me A ReportSkill
- Estimate My BASSkill
- Extract DocumentSkill
- Find My Next MoveSkill
- Five Day ChallengeSkill
- Follow-up RadarSkill
- Form RadarSkill
- Grill Me On This DecisionSkill
- import-from-anywhereSkill
- lead-triageSkill
- Learn My BusinessSkill
- Lint DocumentSkill
- Lint FormSkill
- Lint Nurture SequenceSkill
- log-this-callSkill
- make-it-happenSkill
- Make My VideoSkill
- Make Product DemoSkill
- Make Social PostSkill
- Make ThumbnailSkill
- missed-call-recoverySkill
- Nurture HealthSkill
- Onboard Team MemberSkill
- Outstanding DocumentsSkill
- Outstanding InvoicesSkill
- Package My VideoSkill
- Plan My AdsSkill
- Plan My ContentSkill
- Plan My RoadmapSkill
- Plan My YouTubeSkill
- Prep For CallSkill
- Price My WorkSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 44
Before you fetch anything, know:
- Distinctive: line 52
No website?: Don't stall. Run the local-visibility pass (Step 2) on its own, and make "a simple, findable one-page site" one of the fixes.
- Very specific: line 64
If a fetch is slow, blocked, or empty, say so and offer the fallback: "paste me what's on that page and I'll read it the same way," or fall …
- Very specific: line 101
Only if the owner wants it, offer to read one named local competitor for coverage they have and the owner doesn't.
≈ “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
Get Found Online
You hand the owner a single sharp read: are they findable when someone searches for what they do, and the few fixes that will move that most, worst-first, each with the exact change to make. It reads like a switched-on operator audited them in ten minutes, not like an SEO tool's raw report.
This skill reads the live web. It needs a connection to pull the owner's
pages and their local search presence. No accounts, no key: it works on day one.
The method and the full on-page checklist live in
knowledge/seo-method.md; the local fix order
lives in business-method.md §10.5 (the
local gravity stack). Read both before you order a fix list.
Keyless scope (HARD), per knowledge/research-method.md:
use Firecrawl scrape and search only. Never crawl, map, agent, or
extract (paid, off-floor). Read link and schema signals from a page you already
scraped; never follow them into extra fetches. This per-page-scrape model IS the
bound.
Step 0 — Anchor on this owner (bounded)
Before you fetch anything, know:
- Who the owner is — trade, patch, who they serve. Pull it from the business profile if one exists; if not, ask in one line.
- The site + the key pages — the site URL and the ~3-5 pages that matter (home + top services/contact). Hard cap: ~5-8 page-scrapes. If they name more, audit the most important and say so plainly (never silently drop pages).
- Business name + location — for the local-visibility search.
No website? Don't stall. Run the local-visibility pass (Step 2) on its own, and make "a simple, findable one-page site" one of the fixes. A business with no site can still win the local pack.
Step 1 — On-page pass (scrape each key page)
scrape each named page, one at a time, up to the cap. For each, read from the
scraped content only (never follow its links out): title tag, meta description,
single H1, heading structure, image alt text, internal links, schema hints, and
for local a visible NAP + one clear CTA. Score against the on-page checklist in
seo-method.md and note the specific misses per page.
If a fetch is slow, blocked, or empty, say so and offer the fallback: "paste me
what's on that page and I'll read it the same way," or fall back to the built-in
WebSearch/WebFetch. Never guess at what's on a page you couldn't read.
Step 2 — Local-visibility pass (search)
search the business name, and "[main service] [suburb]". Read: does the
business show up at all; is there a Google Business Profile presence (reviews,
recency, categories, photos, hours); is the name/address/phone consistent with
the site. This is usually the highest-leverage area for a local business.
Order every local fix by the §10.5 gravity stack (1 answer speed → 2 review engine → 3 profile completeness → 4 proof publishing → 5 community presence → 6 paid local), and honor its hard gate: do not lead the fix list with content or keyword work while answer-speed or a review-ask is missing — those come first, because they win more local jobs than any ranking tweak.
When the owner's stated goal is ranking/keyword work but the gate isn't clear, never silently reorder — that reads as "you ignored what I asked" at the exact moment they're leaning in on their one ask. Validate the ask first, then bridge, then present the reordered list: "Chasing 'plumber [town]' makes total sense, it's what everyone says. Before we spend a word on it though, two things ahead of it will win you more jobs faster, and here's why…" Name why the reorder serves their goal (a ranking that sends callers to a voicemail or a stale profile pays to lose the job at the last step). The owner should feel helped toward their goal, never overridden.
Step 3 — SERP / winnability spot-check (search)
For 1-3 target terms the owner would want to win, search and read the live top
results the way seo-method.md describes: what page-type wins (map pack,
directories, a competitor page), is the term realistically winnable, and what the
ranking page does that the owner's doesn't. This grounds any topic advice in
reality instead of a volume number. Keep it to 1-3 terms.
Step 4 — Competitor content-gap (owner-invited, one rival)
Only if the owner wants it, offer to read one named local competitor for coverage
they have and the owner doesn't. Delegate to research-a-competitor (its
search lens) rather than re-implement the read here. One rival, one page.
Step 5 — The prioritized fix list (the deliverable)
Turn everything into 3-7 fixes, highest-leverage first, ordered by the gravity stack for a local business (§4 item 7: a diagnosis outputs a few real moves, not eight pages, so lead with the top ones). This is NOT a keyword laundry list.
Each fix names three things:
- What to do — in plain words.
- Why it moves the needle — the result it unlocks.
- The exact change — the rewritten title tag written out, the review-ask script, the one page to add, the NAP to make consistent. Usable today.
Show the list inline. Offer to save it to seo/get-found-online-audit.md if they
want it on file.
Step 6 — The connected doorway (reactive, outcome-only)
Close by naming what gets sharper when an SEO tool is connected, as outcomes, not
a pitch: real search volumes and difficulty (vs today's winnability read), a
backlink profile, ongoing rank tracking over time, and AI-visibility at scale.
Surface this the BOS way: through what's-now-possible, never a cold CRM pitch.
This is the standard connect-doorway shape (keyless win first, enhancement as an
outcome, the connection that unlocks it), articulated once in
knowledge/connectors.md, referenced here, not re-derived.
Everything here works with zero accounts. When you're ready, connecting an SEO tool lets me track your positions over time and see the real search demand, so we'd know exactly which terms are worth the effort. Totally your call.
Output shape
The audit and fix list are the customer-facing deliverable. Each fix names the exact change to make and the result it unlocks, written so it's usable today.
Hard rules
- Content guardrails. Customer-facing copy uses no em dashes, invents no
facts, quotes, or numbers, and names no third-party vendor. Write it in the
owner's brand voice; the framing and marketing psychology are the owner's
choice. The rules are in
knowledge/content-rules.md. The owner's voice lives inmarketing-strategy/<BrandName>/voice.mdwhen it exists; say so plainly if it does not. - ❌ No
crawl/map/agent/extract. Per-pagescrape+searchonly, capped at ~5-8 pages. Those paid ops are not keyless and not on the floor. - ❌ Don't invent data. No made-up search volumes, no invented review counts, no ranking numbers you can't see. "No public reviews found" is a real finding.
- ❌ Don't lead a local fix list with keyword/content work while answer-speed or a review-ask is missing (§10.5 hard gate).
- ✅ Confirm the site/business is the right one before leaning on the read.
- ✅ Bound it: ~5-8 pages, 1 competitor (invited), 1-3 terms. Finishable in one sitting.
- ✅ If a fetch fails, say so and offer to read pasted content instead.
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.

