Production-ready n8n workflow templates that turn CoreClaw's Google Maps scraper into closed business loops: scrape → score → write to Google Sheets → email a summary with an Excel attachment.

Built for n8n-nodes-coreclaw 0.5.1 (node description.version: 2, 39 operations on /api/v2/). Verified end-to-end on n8n 2.30.4.

What's here

Workflow	What it does	Run mode	Downstream
workflows/gmaps-leads-to-sheets.json	Run the Google Maps scraper, flatten + score rows, append to a Google Sheet.	Run and Get Results (poll)	Google Sheets
workflows/gmaps-leads-sheets-email-summary.json	Same, plus export the run as .xlsx, download it, and email an HTML summary (Top-10 table + email-coverage stats) with the xlsx attached.	Run and Get Results (poll)	Google Sheets + Gmail
workflows/gmaps-leads-callback-export.json	Event-driven: run async with callback_url, and on CoreClaw's callback fetch results, export xlsx, email it. No polling — needs n8n reachable from the internet.	callback webhook	Gmail

Instagram (scrape → score/sentiment → Sheets, with optional email + Excel):

Workflow	What it does	Run mode	Downstream
workflows/instagram-reels-to-sheets.json	Scrape reel data by Reel URL, score each by engagement (likes + comments + plays on a viral-reel scale), append to a Google Sheet.	Run and Get Results (poll)	Google Sheets
workflows/instagram-posts-sheets-email.json	Scrape post data by URL, score by engagement, append to Sheets, and email an HTML Top-10 table + coverage stats with the full dataset attached as .xlsx.	Run and Get Results (poll)	Google Sheets + Gmail
workflows/instagram-comments-sheets-email.json	Scrape comments from posts/reels, classify each by sentiment (heuristic keyword + emoji match), append to Sheets, and email a Top-10 table + sentiment breakdown with the dataset as .xlsx.	Run and Get Results (poll)	Google Sheets + Gmail

Each group is a deliberate complexity ramp: pick the smallest one that does the job.

Use as an n8n template source (one-click import)

These three workflows are published as a self-hosted n8n template feed so anyone can import them from the in-app Templates panel — no manual JSON download. Point your self-hosted n8n at the live Worker:

N8N_TEMPLATES_HOST=https://coreclaw-n8n-templates.t445481611.workers.dev

Restart n8n, open the Templates panel, and the three CoreClaw workflows appear under Sales → Lead Generation. Click → Use workflow → it lands in your canvas (credentials stripped — you bind your own on import).

The Worker source and deploy instructions live in template-host/. It implements the full N8N_TEMPLATES_HOST contract (categories, search, detail, flat import, collections) verified against api.n8n.io. Deploy your own fork with wrangler deploy.

Why this replaced the old pack

The previous 12 workflows in legacy-v1/ are archived because:

1. They targeted the v1 node contract (scraperSlug + scraper.run/run.get/run.getResults + explicit version: v1.x.x). The v2 node uses worker_id + Run and Get Results; the parameters no longer match.
2. They claimed Google Sheets / Gmail / Airtable integration but contained zero such nodes — only Code nodes that assembled payload field names. These workflows actually wire the downstream nodes.
3. 9 of 12 were the same Google Maps scraper with different post-processing, generated from a single source script. These three are hand-written against the real worker schema.

Prerequisites

• n8n 2.22.5+ (built and import-tested on 2.30.4).
• The n8n-nodes-coreclaw community package installed (Settings → Community nodes → install n8n-nodes-coreclaw). On self-hosted Docker you can also npm install n8n-nodes-coreclaw inside the n8n user folder and restart.
• Three credentials — see docs/credentials-binding.md and docs/google-oauth-setup.md:
  • CoreClaw API (key + base URL https://openapi.coreclaw.com)
  • Google Sheets OAuth2
  • Gmail OAuth2 (workflows B and C only)

Quick start

1. Install the node package and create the credentials above.
2. Create a Google Spreadsheet with a sheet named Leads (workflows A and B write there).
3. Import a workflow JSON: n8n → Workflows → ⋮ → Import from File.
4. Bind credentials on every node flagged REPLACE_WITH_* (see docs/credentials-binding.md).
5. For workflow C, set callback_url on the Start Run node to your CoreClaw Trigger webhook URL — and make sure n8n is reachable from the public internet (or via a tunnel like cpolar/frp).
6. Open the Input Config node, set your keywords / base_location / max_results, click Execute workflow.

The default search (HVAC Contractors / New York, USA / max_results=5) costs roughly $0.008 per run at the worker's published rate and finishes in ~30–40s.

How the scoring works

The Normalize & Score Code node computes lead_score (0–100):

score = round(min(100, review_rating · log10(review_count + 1) · 12  +  hasEmail·15  +  hasWebsite·5))

Tune the weights in the Code node. Rows are sorted by lead_score descending before they hit the sheet and the email table.

Field mapping

Worker result fields → sheet columns are documented in docs/field-map.md. That file also lists the four places the worker's README sample disagrees with real output (e.g. review_rating/review_count, not rating/reviews_total; all_emails is a string, not an array) — these workflows use the real fields.

Validation

Both polling workflows were executed end-to-end against a live CoreClaw account and a real Google account on n8n 2.30.4 (worker coreclaw/google-maps-scraper, search HVAC Contractors / New York, USA / max_results=5):

Workflow	Result
A — gmaps-leads-to-sheets	✅ success — 5 records scraped, scored, appended to a Google Sheet (Leads tab, 20 columns).
B — gmaps-leads-sheets-email-summary	✅ success — same scrape + Sheet append, plus .xlsx exported (64 KB) and emailed to the configured recipient with an HTML Top-10 summary.
C — gmaps-leads-callback-export	Imported and activated; runtime requires a public/tunnel webhook URL (not exercised locally). See the workflow's sticky notes.

Instagram workflows were also executed end-to-end (worker coreclaw/instagram-post-scraper and instagram-comment-scraper, post https://www.instagram.com/p/DZ5T2XPllXv/):

Workflow	Result
D — instagram-reels-to-sheets	✅ success — reel scraped by URL (worker coreclaw/instagram-reel-scraper), scored (engagement_score 62), appended to the Reels sheet (3.49M likes reel).
E — instagram-posts-sheets-email	✅ success — post scraped + scored (engagement_score 28), appended to the Posts sheet, .xlsx exported (12.6 KB) and emailed with an HTML Top-10 summary.
F — instagram-comments-sheets-email	✅ success — 20 comments scraped + sentiment-classified (5 positive / rest neutral/negative), appended to the Comments sheet, .xlsx exported (239 KB) and emailed with a sentiment breakdown.

Google Sheet created by the runs: a spreadsheet titled Instagram Insights with Reels, Posts, and Comments tabs (19/18/11 column headers respectively).

Google Sheet created by the run: a spreadsheet titled CoreClaw Maps Leads with a Leads sheet and a 20-column header row (lead_score, search_rank, title, phone, website, emails, primary_category, categories, review_rating, review_count, status, city, state, address, latitude, longitude, place_url, source_keyword, source_location, scraped_at).

Node-package bug (fixed in 0.4.1)

During validation, the Run and Get Results node failed with NodeOperationError: Could not find property from collectParams. Root cause: in n8n-nodes-coreclaw ≤ 0.4.0, the runAndGetResults/rerunAndGetResults specs spread runBodyParams (callback_url / is_async / body offset+limit) and resultPaginationParams into spec.params, but the node description only displays those fields for run/rerunLastRun/abortLastRun — not for runAndGetResults. findDisplayedProperty therefore threw.

Fixed upstream in 0.4.1 (commit): the composite specs now keep only workerId/version/input_json/raw_input_json in spec.params (returnAll is fetched separately in router.js; offset/limit fall back to 0/50). The fix is published to npm, so anyone installing n8n-nodes-coreclaw@latest (≥ 0.4.1) gets the working version — no patching needed.

Troubleshooting

Symptom	Cause / fix
Node shows REPLACE_WITH_* / red flag	Bind the credential (see credentials-binding.md).
CoreClaw 12001 / 12002	Invalid or unauthorized API key — recreate the CoreClaw API credential.
CoreClaw 11004	Worker not found — the workerId mode id value must be coreclaw~google-maps-scraper (owner path) or the store slug.
CoreClaw 30001	Insufficient balance on your CoreClaw account.
Run did not finish before polling timed out	The Run and Get Results node polls ~4 min max. For long jobs, use workflow C (callback) instead of polling.
Google redirect_uri_mismatch	OAuth redirect URI must be exactly http://localhost:5678/rest/oauth2-credential/callback (see google-oauth-setup.md).
Gmail invalid_scope	Use gmail.send only.
Workflow C never fires	n8n isn't reachable from the internet; CoreClaw can't POST to your webhook. Use a tunnel or deploy n8n publicly.
Empty emails / categories columns	You're on a worker version whose output still uses arrays — check docs/field-map.md "Gotchas".

Repository layout

.
├── workflows/          # the three v2 workflow JSON files (import these)
├── docs/               # credential setup, binding guide, field map
├── legacy-v1/          # archived v1 workflows (do not use with the v2 node)
└── README.md

License

MIT — same as the upstream node package.

---