Publish SEO content that ranks

Content lead / SEO / solo founder / head of growth · Assemble in ~20 min; first published article from keyword to live page in under 2 hours per batch.

GTMEdit flowon X◆ new (unverified) 90

step 1ready

Keyword research and priority plan

seo-strategist3 alts

step 2ready

Content brief per target keyword

seo-content-brief15 alts

step 3ready

Full draft in brand voice

write-content28 alts

step 4ready

On-page SEO + GEO / AEO optimization

seo-aeo-amplifier29 alts

step 5partial

Publish to CMS

seo-page-builder15 alts

step 6partial

Monitor rankings and recover drops

on-page-seo-auditor29 alts

Did this recipe work for you?

What you get

▸Prioritized keyword roadmap with winnability scores — reusable across batches
▸Research-backed content briefs (one per target keyword)
▸Published, GEO-optimized articles with JSON-LD schema — live on your CMS
▸Per-page audit reports with prioritized recovery actions for any rank drops

tools: Google Search Console (free) for rank data · CMS hosting (existing) · optional rank-tracker for scheduling (Ahrefs/Semrush from $99/mo or free GSC export)

tokens: ~$2-6 per article end-to-end (brief + draft + optimize)

≈ $10-25/mo in tokens at 5-10 articles/week; rank-tracker optional add-on cost depends on your existing stack

What stays yours

· Approve briefs and drafts before publishing — brand voice alignment and factual accuracy require a human review gate
· Supply and maintain CMS API credentials; adapt the publish step for non-WordPress platforms
· Interpret audit reports and decide which recovery actions to prioritize based on business context

The problem

When I need a steady stream of ranking content but have no writing team, I want an agent to run the full lifecycle — keyword research, brief, full draft in my brand voice, on-page + GEO optimization, publish to CMS, and monitor/recover — so I get agency-tier organic output in-house instead of a $5K-$15K/mo deliverable-count retainer.

Cybernetic layers

Memorybyo

Published-article registry (URL, target keyword, publish date, last rank check). No harvested memory artifact yet — use a Google Sheet, Notion database, or a local markdown file. Required to avoid re-briefs and to route monitoring back to the right page.

Background-schedulingbyo

Weekly rank-check loop (step 6 should run on a schedule). A Claude Code cron task or external scheduler (GitHub Actions, cron job) wired to the article registry enables unattended monitoring. No library artifact covers this yet — it is the primary gap.

Harnessready

Claude Code is the runtime. Drop each ready skill into ~/.claude/skills/ and invoke by name. Steps 1-4 chain cleanly in one session; steps 5-6 may need separate invocations per CMS credentials and monitoring cadence.

Click any step to see what fills it — and the ranked alternatives you can swap in.