Just dropped a walkthrough of the Local SEO Brain - a source-cited Obsidian vault + Claude Code agent that ranks any local business in the Google Map Pack, organic local SERPs, and AI search. If you're running local SEO (yours or clients'), this replaces the spreadsheet-and-screenshot grind. Setup ~30 minutes per business, then the agent handles audits, GBP optimization, backlinks, and review strategy in a loop. Works with Claude Code, Codex, or Gemini. Built on Andrej Karpathy's Hot/Index/Wiki memory pattern so the vault grows with every session. Watch the full walkthrough: https://youtu.be/4v7116sEbrQ Want the vault? it is here: https://skool.com/ai-marketing-hub-pro

Hey everyone, just shipped claude-blog v1.9.1 What changed: - 5 delivery gates now run before any draft reaches you (capability, format, visual, BLOCKING review, integrity) - 8-agent cybersecurity audit closed all HIGH and MEDIUM findings - 187 tests, score lifted to 95 of 100, Exceptional band - Image fallback ladder: Nano Banana → Gemini → premium stock → Openverse → block-with-setup (no more silent 404s) - Custom rules layer: pass things like "no em-dashes" or "use organic tone this week" and the reviewer scores against them too - AI Overviews + AI Mode optimization built into every draft (answer-first takeaways, sourced stats, FAQ blocks) The big shift: the user is no longer the first reviewer. The gates are. Walkthrough (10 min, agent builds a real Codex vs Claude blog live): https://www.youtube.com/watch?v=7Q4GaSgUFHo Repo: https://github.com/AgriciDaniel/claude-blog

Hey hub. Just dropped Claude ads 1.7. The biggest release since we started. What changed: - /ads amazon is in. Sponsored Products, Brands, Display. ACOS, TACOS, search-term harvesting, brand-defense bidding. $56B TAM, zero existing skill coverage before this. - Google AI Max pre-flight. Sept 2026 your DSA, ACA, and campaign-level broad-match get auto-migrated. claude-ads catches every legacy campaign first so you set the migration order, not Google. - Meta Entity-ID predictor. Andromeda is throttling near-duplicate creative before delivery. We score it on five axes pre-launch, not post. - Cross-host install: same skill, six hosts. Claude Code, Codex CLI, Cursor, Windsurf, Gemini CLI, Goose. - Eval harness: 41 pytest tests enforce a 209-row check catalog. Same audit twice equals the same grade. CI fails on drift. Watch the walkthrough: https://youtu.be/qUXSZHiMPK8 Install: github.com/AgriciDaniel/claude-ads Talk to me in the comments: what platform should /ads cover next? Wave 3 is Walmart Connect and Connected TV. Drop your specific use case and I will shape the audit checks around what you actually run.