Everything auxfirst publishes for ad holdings, media agencies, brands, publishers, and platforms. The ebook, the indexes, the headless protocol guide, the agency-brains framework, and the carousel deck. One trust thesis, six artefacts. Built for the agency floor, not the demo stage.
The ebook is the conceptual frame the rest sits on. Read this first if you have not. The trust layer, the design patterns, the autonomy model, the escalation paths, the memory policies, the audit trails — everything you need to turn an agent idea into a service a global client will actually sign off on.
The 10 heuristics, the 4-stage trust architecture, the autonomy mapping, the 90-day roadmap, the 10-dimension scorecard. The artefact every other piece on this hub assumes you have read.
№ 01 maps the cognitive layer — 120 agents inside the agency, brand, and holdco. № 02 maps the transacting layer — 50 specialist agents that face other agents across the new agentic ad-buying world. Read together; they cover the full surface.
A working catalogue of 120 trustworthy agentic services for ad holdings, creative networks, media agencies, commerce shops, CRM, and shopper teams. Each with visible intent, evidence, autonomy boundaries, escalation paths, memory policies, and audit trails.
Open the 120 → 02 50 agents · 5 categoriesThe agents that face other agents. Standards bridges, channel buyers, publisher counter-agents, inline trust & QA, strategic specialists. The companion to № 01, built for the agentic ad-buying world the May 2026 upfronts made real.
Open the 50 → ↻ 120 agents · Apple-style deckThe 120 Index as a full-screen carousel — for client calls, leadership readouts, and quick skimming. Keyboard, swipe, category jump. Not a replacement for the wiki — a different reading surface.
Open the deck →Two foundational pieces an agency reads before building anything. MCP for Advertising answers the headless question — what changes for an agency stack when the Model Context Protocol arrives, and what does not. Agency Brains Design answers the cognition question — which kind of reasoning each agent surface needs, named and governed.
What the Model Context Protocol changes for an agency stack — and what it doesn't. Five structural facts, five topologies, four maturity tiers. The plug is the easy part. The fiduciary boundary is the hard one.
Read the guide → F1 Specification · Taxonomy · ManifestThe working framework for defining, naming, and governing the AI brains an ad agency runs. The taxonomy, the meta-brain router, the manifest format, the agency-specific stack, and the engagement that builds it.
Read the spec → F2 Implementation Atlas · Turn-by-turnThe build sequence. Applied patterns, worked brain examples, and the leg-by-leg, turn-by-turn engagement guide for ad agencies adopting the Brains framework. Pairs with № 01.
Open the atlas →If you are landing here for the first time, this is the order. Each step assumes the one before it. The whole sequence takes a long afternoon. You can also commission us at any step — most agencies do, somewhere between step three and step five.
auxfirst is an agentic experience design agency. We don't license agents and we don't resell platforms. We design the trust layer — the visible intent, the evidence, the autonomy boundaries, the escalation paths, the memory policies, the audit trails — that turns one of these artefacts into a service a holdco will put in front of a global client.