[TERMINAL · SKILLS]
> mounting /skills...
> indexing 295 manifests...
> linking agents: claude · codex · gemini · cursor
> ready.
[░░░░░░░░░░░░░░░░░░░░░░░░░░░░] 0%
Terminal.skills
Use Cases/Automate Local SEO Monitoring for Client Businesses

Automate Local SEO Monitoring for Client Businesses

Wire an AI agent to the SEOG MCP server to onboard client businesses, track map-pack rankings by neighborhood, triage and draft review replies, watch nearby competitors, and produce a weekly local-SEO digest — no dashboard babysitting.

Business#local-seo#google-business-profile#mcp#rank-tracking#reviews
Works with:claude-codeopenai-codexgemini-clicursor

Skills stack · 3 skills

Avg quality 93/100·All SAFE
>

seog

v1.0.0

Manage local SEO from your terminal via the SEOG MCP server — Google Business Profile businesses, keyword rank tracking, review monitoring and reply drafts, and competitor intelligence. Use when: connecting an agent to seog.ai, tracking map-pack rankings, automating local SEO reports, monitoring competitors near a business, drafting review responses, or when the user mentions "SEOG", "local SEO", "map pack", "Google Business Profile rankings", or "keyword positions" for a physical business.

93/100 quality
10.00× impact
SAFE
View skill
>

seo-audit

v1.0.0

When the user wants to audit, review, or diagnose SEO issues on their site. Also use when the user mentions "SEO audit," "technical SEO," "why am I not ranking," "SEO issues," "on-page SEO," "meta tags review," or "SEO health check." For building pages at scale to target keywords, see programmatic-seo. For adding structured data, see schema-markup.

93/100 quality
1.50× impact
SAFE
View skill
>

schema-markup

v1.0.0

When the user wants to add, fix, or optimize schema markup and structured data on their site. Also use when the user mentions "schema markup," "structured data," "JSON-LD," "rich snippets," "schema.org," "FAQ schema," "product schema," "review schema," or "breadcrumb schema." For broader SEO issues, see seo-audit.

93/100 quality
2.91× impact
SAFE
View skill
$

The Problem

Dana runs a 4-person marketing agency with 15 local clients — dentists, cafés, a med spa. Every Monday someone burns half a day clicking through Google Business Profiles: checking where each client ranks for their money keywords, which reviews came in over the weekend (and which angry ones need an answer today), and whether the competitor across the street overtook them. It's repetitive, error-prone, and unbillable. Rankings are checked from the office — not from the neighborhoods customers actually search from — and nobody notices a rival's review surge until the client does.

The Solution

Connect your coding agent (Claude Code, Cursor, Gemini CLI) to the SEOG MCP server. SEOG tracks map-pack rankings, reviews, and competitors for physical businesses; its MCP endpoint exposes all of it as 25 tools, so the Monday ritual becomes one prompt — or a scheduled agent run that lands a digest in Slack.

Step-by-Step Walkthrough

Step 1: Connect the agent to SEOG

Issue a token in seog.ai → Settings → MCP access, then:

bash
claude mcp add --transport http seog https://api.seog.ai/mcp \
  --header "Authorization: Bearer $SEOG_MCP_TOKEN"

Step 2: Onboard each client once

"Import 'Bright Smile Dental, Austin' into SEOG and start tracking its money keywords."

The agent runs search_places("Bright Smile Dental Austin"), imports the right Places result, then keyword_recommendations → add_keyword for the winners — with locationLabel per neighborhood ("Hyde Park", "South Congress") so rankings are measured where patients actually search, not from the office IP.

Step 3: Schedule the weekly sweep

One prompt, run on a schedule (cron, CI, or your agent's scheduler):

"For every business: check keywords, sync reviews, snapshot watched competitors, then write me a digest with rank movement, unanswered negative reviews (draft replies), and competitor alerts."

The agent loops list_businesses → per business: check_keyword on active keywords, sync_reviews + list_reviews(filter="needs-response") + draft_review_response for each (drafts only — the owner approves in-app), snapshot_competitor on watch-listed rivals.

Step 4: Escalate what matters

The digest ranks findings by impact: a client dropping out of the 3-pack for a money keyword beats a 4-star review without a reply. For medical/legal clients the drafted replies stay generic — never confirming a patient visit (review policy).

Real-World Example

A Bratislava café client, week 3 on the system: the agent's sweep found the café at 4.3★/957 reviews while discover_competitors(radius=1000) surfaced La Putika 2 at 4.2★/962 — a review-count race the client was losing by literally five reviews. The digest flagged it, the café ran a two-week table-QR review push, and the agent's keyword_history showed "coffee shop bratislava" climbing #4 → #2 as review velocity recovered. The rival had no website — the digest recommended doubling down on the client's site (see the seo-audit skill) — and set_competitor_watchlist now alerts the day the rival's rating or review count jumps.

Related Skills

  • seog — the MCP integration this workflow runs on (businesses, keywords, reviews, competitors)
  • seo-audit — diagnose the client's website when the map-pack data says the site is the weak signal
  • schema-markup — add LocalBusiness structured data, the most common fix the audits surface for local clients