Skip to content

feat(seo): consolidate six SEO PRs into one - #551

Draft
JayZeeDesign wants to merge 12 commits into
mainfrom
cursor/seo-consolidated-cd57
Draft

JayZeeDesign wants to merge 12 commits into
mainfrom
cursor/seo-consolidated-cd57

Conversation

@JayZeeDesign

Copy link
Copy Markdown
Contributor

Summary

Consolidates six open SEO PRs into a single coherent change set. All conflicts resolved; tests green.

What came from each PR

From #481 (cursor/seo-title-h1-alignment-f93f)

  • /tools/{provider} titles now match H1s
  • Metered: {Provider}: {n} tools from {price} | treg.to
  • Mixed: {Provider}: {n} tools, platform or your own key | treg.to
  • Own-account: {Provider}: connect your own account | treg.to
  • Dropped leftover API pricing chrome
  • Added tests in test_provider_pages.py

From #500 (cursor/ecommerce-usecase-hero-rewrite-cce8)

  • Amazon ASIN and TikTok Shop use-case hero/first-fold rewrite in agent_pages.py
  • Tighter hero copy that leads with the buyer outcome

From #503 (cursor/reddit-x-pain-points-2fcf)

  • Reddit/X pain FAQ + failure_modes on five workflow pages in agent_pages.py
  • Added WORKFLOW_FAILURE_MODES for: find-and-verify-a-lead-list, screen-instagram-creators-before-outreach, discover-creators-in-a-niche, keyword-demand-to-ad-budget, mine-competitor-meta-ads-as-creative-pack

From #511 (cursor/mcp-intent-seo-fix-6dcb)

  • MCP lead-in Title/H1 for google-search-console, google-analytics, semrush own-account pages
  • _MCP_INTENT_PROVIDERS allowlist for MCP-intent titles that win over generic own-account pattern
  • Meta descriptions name MCP plus connect-own-account plus treg.to as one MCP for the catalog
  • Updated docs/context/interface/seo.md title-contract docs

From #544 (cursor/pricing-receipts-53d5)

  • Dated enrichment/SEO/ads/creator receipts on /pricing
  • Cost FAQ items: call cost vs usable-result cost, BYOK billing
  • Figures from real workflow runs (no invented numbers)

From #547 (cursor/work-email-finding-bench-b568)

  • /blog/work-email-finding-bench receipt post
  • Blog index + sitemap + route registration
  • Comprehensive tests in test_blog.py

Conflict resolutions

Supersedes

Test results

255 tests passed (test_seo, test_provider_pages, test_agent_pages, test_blog)

Open in Web Open in Cursor 

cursoragent and others added 12 commits September 14, 2026 00:17
- Remove 'API pricing:' lead-in from metered/mixed provider page titles
- Metered titles now: '{Provider}: {n} tools from {price} | treg.to'
- Mixed titles now: '{Provider}: {n} tools, platform or your own key | treg.to'
- Own-account titles unchanged: '{Provider}: connect your own account | treg.to'
- Update seo.md docs to reflect Title=H1 rule
- Rename test_provider_title_leads_with_pricing -> test_provider_title_matches_h1

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
…pages

Tighten the first fold (H1, lede, title) on two ecommerce use-case pages
to match the converting people-search LP and hub patterns:

/use-cases/amazon-product-detail-by-asin
- H1: 'Pull any Amazon product by ASIN' (was keyword essay)
- Lede: concise buyer outcome, price from catalog, no affiliate required

/use-cases/tiktok-shop-products-and-reviews
- H1: 'Search TikTok Shop products and reviews' (was API-focused)
- Lede: concise buyer outcome, price from catalog, no seller account needed

Both pages now have:
- Clear buyer job in the headline
- Catalog-backed price in title and lede ({cheapest} template)
- Concise copy without em-dashes
- Proper treg.to naming throughout

No new URLs, no fabricated metrics, no changes to comparison tables or FAQ.

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
Append failure_modes and FAQ items sourced from Reddit and X discussions
to the five live workflow pages merged in PR #462:

- screen-instagram-creators-before-outreach: +3 failure_modes, +2 FAQ
- discover-creators-in-a-niche: +3 failure_modes, +2 FAQ
- keyword-demand-to-ad-budget: +3 failure_modes, +2 FAQ
- mine-competitor-meta-ads-as-creative-pack: +3 failure_modes, +1 FAQ
  (skipped duplicate FAQ re official Meta API)
- category-content-intel-tiktok-xiaohongshu: +3 failure_modes, +2 FAQ

Update test assertion from == 4 to >= 4 for workflow FAQ count to
allow appended items.

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
…e-analytics, semrush

GSC shows strong "{provider} mcp" impressions with near-zero clicks on these
own-account /tools/{provider} pages when their titles said only "connect your
own account".

- Add _MCP_INTENT_PROVIDERS allowlist for the three providers
- Title: "{Provider} MCP: connect your own account | treg.to"
- H1: matches Title lead
- Meta description: names MCP + connect-own-account + treg.to as one MCP
- Lede: mentions one MCP for the whole catalog
- Update docs/context/interface/seo.md with the new contract
- Add test_mcp_intent_providers_lead_with_mcp pinning all three

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
Add catalog-backed receipt examples that teach call cost vs usable-result
cost, billing unit, misses, and BYOK. All figures sourced from existing
workflow run receipts (no invented numbers):

1. Enrichment: verified lead list (2026-08-26)
   - 50 companies, $3.62 total, $0.13 per deliverable lead
   - Shows per-success misses at $0.00 vs per-call billing

2. SEO: keyword demand (2026-09-14)
   - 50 keywords, $0.11 total (ideas + volume batch + trend)
   - Demonstrates batch billing (volume call: per-request, not per-keyword)

3. Ad library: Meta and Google ads (2026-09-14)
   - BYOK example: Meta calls on own Apify key (not metered)
   - Google shared-key: $0.015

4. Creator discovery: Instagram (2026-09-14)
   - 25 creators, $0.20 total
   - Per-creator discovery billing

New FAQ items:
- What is the difference between call cost and usable-result cost?
- Are own-key calls billed?

Source workflow pages:
- /workflows/find-and-verify-a-lead-list
- /workflows/keyword-demand-to-ad-budget
- /workflows/mine-competitor-meta-ads-as-creative-pack
- /workflows/discover-creators-in-a-niche

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
Receipt post comparing work-email finding across vendors based on a
2026-09-16 neutral run of 292 people with published team-page emails.

Key findings:
- Quality tie: treg, Clay, Freckle at 262-264 exact matches
- Cost gap: treg $0.0056/correct vs Clay $0.0395/correct (7x)
- Fair Clay framing included for visual/GTM/spreadsheet workflows

Includes:
- /blog/work-email-finding-bench route with full data table
- Disclosures: list bias, MillionVerifier re-verify, 502 bug
- Links to /pricing, /people-search, /workflows, /use-cases
- Blog index registration, sitemap entry, route ownership
- 15 new tests covering content, SEO, and constraints

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
…op use-case pages

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
…orkflow pages

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
…nsole, google-analytics, semrush

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
…n /pricing

Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
Co-authored-by: Jason Zhou <JayZeeDesign@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

area:docs Documentation & design fragments

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants