Summary — 2026-05-20
Second pass — 2026-05-20 (three items scored from mobile chat)
Three items surfaced in Anthony's mobile-Claude chat 2026-05-19/20. Mobile-Sebastian pre-scored them without canon access. Real Sebastian re-scores below with full canon (failures.md class-match, brand-fingerprint risk, sellability into Articulate's MEP and the data-sovereign bundle). Mobile scores in parens for delta visibility.
| Item | Source | P | Q | S | Total | Move |
|---|---|---|---|---|---|---|
Hallmark (@nutlope, github.com/Nutlope/hallmark) — anti-AI-slop design skill, npx-installable. 22 themes, 65 slop-test gates. Hassan El Mghari (Vercel; roomGPT 2.3m users, restorephotos 500k). |
Anthony's mobile chat 2026-05-19 | 2 (was 3) | 3 (was 3) | 1 (was 3) | 6/9 (was 9/9) | Act on, but in TripleD's lane. Install only after Belinda canon check vs Articulate brand v3 — Hallmark's 22 themes could conflict with the deployed serif + red + Work Sans. TripleD owns the install + delta-against-canon pass. Mobile-Seb's 3/3/3 was over-eager: brand-spec-drift class-match in failures.md 2026-05-18 (Takis four-brand cycle). Hallmark is not a standalone Articulate product (S=1, not 3) — it's an input to TripleD's existing workflow. |
Google-Agent (Google docs update, confirmed launch 2026-03-20) — new user-triggered fetcher for AI agents browsing on behalf of users. Distinct from Googlebot. Own IP-range file user-triggered-agents.json. Web Bot Auth-signed identity https://agent.bot.goog. |
Anthony's mobile chat 2026-05-19 | 2 (was 3) | 2 (was 2) | 2 (was 1) | 6/9 (was 6/9) | Act on as a billable line item, not just hygiene. Mobile-Seb scored S=1 ("infrastructure hygiene"). Re-score S=2: "AI-agent-readiness audit" maps to a real deliverable — 30-min log + WAF + robots-policy check per property, $300–500 line item inside the data-sovereign-stack bundle. Add to MEP delivery #4 (Lead capture + CRM hygiene) as sub-step. Build the playbook in SuperSebastian/playbooks/google-agent-audit.md (T+7). |
Corey Haines marketing-psychology skill (github.com/coreyhaines31/marketingskills) — Claude Code skill. Anchoring, loss aversion, decoy, charm pricing, AIDA, BJ Fogg, nudge. Part of 23-skill marketingskills repo (CRO, copy, SEO, analytics, growth). npx skills add coreyhaines31/marketingskills. |
Anthony's mobile chat 2026-05-19 | 1 (was 2) | 2 (was 1) | 3 (was 2) | 6/9 (was 5/9) | Act on — but the move is "build the superior Articulate version," not "install Corey's." Mobile-Seb scored S=2 conservatively. Real Sebastian: S=3 because Anthony's edge (Ogilvy/Clicksco/MyAcquisitionExpert/25 years) anchored to Rory Sutherland + Cialdini + Mollick + Fogg becomes a productised Articulate offer surface — the "Marketing Psychology Toolbox." Corey's skill is the benchmark, not the destination. Anthony's offer hosts in Articulate/offers/marketing-psychology-toolbox/. See SuperSebastian/decisions.md 2026-05-20 + new offer doc. |
Bundle implication: Google-Agent audit folds into the data-sovereign-stack bundle as a $300–500 per-property line item. Marketing Psychology Toolbox is stand-alone Articulate offer surface, separate from MEP — sold to clients with an existing marketing function who want their pages audited by a UAE-veteran Ogilvy alumnus running a structured toolkit, not a freelancer with a checklist.
Class-match against failures.md: - Hallmark: 2026-05-18 Takis four-brand-cycle (brand-spec drift). Hallmark install must run through Belinda before any Articulate surface or client deliverable. - Google-Agent: no direct class match. Closest is the post-deploy probe discipline pattern. - Corey Haines: no class match. Net new sellable surface.
Re-score triggers: - Hallmark: after TripleD's first delta-against-canon pass. If quality lifts without breaking Belinda v3, P promotes to 3 (total 7/9). - Google-Agent: when first paid Google-Agent-readiness audit lands, OR Google-Agent volume crosses 1% of Googlebot in any client's logs. - Marketing Psychology Toolbox: when first paying client signs the offer (proves S=3). If 90 days pass with no buyer, S re-scores to 2.