Research note · April 2026
Use cases for the agentic web
Nine briefs—one per profile prefix (`/u` through `/x`). Each ties a real vertical to the matching entity surface in more.md and cites third-party sources from our April 2026 baseline report on agentic verticals.
Why this page exists
Work is shifting from "humans click in browsers" to "agents run tools and move data." Analyst estimates for the agentic AI market land around $7.8B in 2025 and roughly $52.6B by 2030, with a CAGR in the mid-40% range. Gartner's public line is that a large share of enterprise apps could embed task-specific agents by late 2026, up from single digits in 2025. Treat those figures as directional: vendors disagree on definitions, but nobody argues the curve is flat.
The painful part is infrastructure. Layout-heavy HTML and bespoke APIs waste context. People call that cost the "HTML Tax." It shows up as worse answers, higher inference bills, and integrations that break when a template changes.
more.md is meant to sit in that gap: predictable entity URLs, DIDs, Markdown/JSON/TOON profiles, programmable gates, SSE for push updates, and Pulse (WebSocket) for negotiation-shaped traffic. You can use the hosted stack or run it on-prem when PHI, grid gear, or residency rules say you must.
Getting to agentic-web ready on more.md
You do not need a manifesto to ship. You need stable handles, honest formats, and gates that machines can satisfy. A practical readiness checklist looks like this:
- Entity model: nine URL prefixes (`/u` person, `/o` organization, `/l` place, `/a` asset, `/p` product, `/k` knowledge, `/agent` AI agent, `/e` event, `/x` custom) so callers resolve the right profile shape.
- Gates on the Go data plane: payment, trust, identity, credential, connection, capability, combined — with 402/403 responses that tell an agent what proof to fetch next.
- Formats: content negotiation for Markdown, JSON, and TOON; TOON matters when JSON key repetition is eating your token budget on tool calls.
- Real time: SSE from the data plane; Pulse for commerce-style state machines and multi-party coordination.
- SDKs: `@more-md/sdk` (npm) and `more-md` (Python) for the common read/subscribe/search paths.
Pick one sector below, ship a thin vertical on top of gates + SSE + TOON, and measure latency or token cost against your baseline. A buyer who admits the status quo is expensive matters more than polishing protocol trivia in a vacuum.
Potential entry points
Each card maps one vertical to a single profile prefix (`/u`–`/x`). Friction is loud enough in these pockets that a focused build on gates, serialization, and real-time state can still clear a bar even if you keep scope narrow and skip hero-slogan bets.
Person · /u/
HR and hiring
- Market signal
- AI in HR toward ~$15.24B by 2030 (forecast).
- Friction
- Candidates distrust AI scoring; resumes are easy to fake.
- more.md fit
- Primary `/u` candidate profiles; W3C VCs; trust gates; Markdown/TOON for ATS agents.
View sector brief →
Organization · /o/
Procurement and supply chain
- Market signal
- AI sourcing cited as a large wedge of B2B procurement (~$184B segment in one 2025 analysis).
- Friction
- Micro-purchases stall on coordination; supplier data stays messy.
- more.md fit
- Primary `/o` trading partners; linked `/p` SKUs; discovery manifest; Pulse; x402-style settlement.
View sector brief →
Place · /l/
Commercial real estate
- Market signal
- PropTech still draws multi-billion annual capital flows (industry lists).
- Friction
- Physical faults still route through people and tickets.
- more.md fit
- Primary `/l` building and campus profiles; smart sections; SSE for IoT events.
View sector brief →
Asset · /a/
Distributed energy and IoT
- Market signal
- Load growth (data centers, electrification) and edge storage in focus per IEA and industry surveys.
- Friction
- Central dispatch cannot optimize every local spike in time.
- more.md fit
- Primary `/a` for BESS and large loads; on-prem edge; event bus; Pulse + x402 for trades.
View sector brief →
Product · /p/
Retail and e-commerce
- Market signal
- McKinsey discusses agentic commerce on the order of trillions by 2030.
- Friction
- Assistants need comparable catalogs without walled-garden checkout tax.
- more.md fit
- Primary `/p` SKUs and offers; merged Markdown catalogs; HTTP 402 payment gates.
View sector brief →
Knowledge base · /k/
Enterprise knowledge management
- Market signal
- Knowledge-management tech spend in the billions (industry lists).
- Friction
- RAG stacks ship fat JSON and bleed tokens on every tool call.
- more.md fit
- Primary `/k` knowledge bases; doc ingest to Markdown; `Accept: text/toon`; MCP for IDE agents.
View sector brief →
AI agent · /agent/
Healthcare and life sciences
- Market signal
- North American agentic healthcare AI ~$0.66B (2025).
- Friction
- HIPAA, liability, and handoffs between clinical and admin agents.
- more.md fit
- Primary `/agent` clinical & ops automations; `/o` care orgs; verifiable-credential gates; on-prem.
View sector brief →
Event · /e/
Corporate travel and events
- Market signal
- Tourism AI spend projected toward ~$13.38B by 2030.
- Friction
- Complex B2B trips and RFPs still live in email.
- more.md fit
- Primary `/e` for conferences, holds, trip milestones; Pulse sequencing; linked `/o`/`/l` in the graph.
View sector brief →
Custom · /x/
Legal and enterprise governance
- Market signal
- AI governance software globally in the mid–hundreds of millions; EU AI Act fines can reach tens of millions of euros in worst cases.
- Friction
- Per-action authorization trails; vague human-in-the-loop claims fall apart under audit.
- more.md fit
- Primary `/x` for jurisdiction-specific assurance payloads; changelogs; DIDs; capability gates.
View sector brief →