Canonical Resources
A curated source list for tracking the headless economy — protocols, payments, AI platforms, discovery standards, hybrid companies, and analysis streams.
A curated source list for tracking the headless economy: the shift from GUI-first software toward machine-native surfaces that agents can discover, evaluate, access, pay for, and operate through protocols, APIs, CLIs, SDKs, MCP servers, structured content, and agent-facing documentation.
| Resource | Description |
|---|---|
| Protocols and standards | |
| Model Context Protocol | The canonical protocol for connecting AI agents and applications to tools, data, APIs, and external systems. |
| Agent2Agent Protocol | Open protocol for agent-to-agent discovery and communication, including the Agent Card pattern for publishing agent capabilities. |
| Agent Communication Protocol | Open standard from IBM Research and the Linux Foundation for agent-to-agent messaging, task hand-off, and lifecycle. Layers above MCP — where MCP exposes tools to a single agent, ACP coordinates between agents across frameworks and runtimes. |
| OpenAPI Specification | Mature standard for describing HTTP APIs in a machine-readable way. Still one of the most important foundations for agent-usable software. |
| agents.json | Emerging proposal for exposing agent-facing API and interaction metadata. Relevant for making products more legible to autonomous systems. |
| llms.txt | Proposed convention for publishing a Markdown entry point that helps LLMs and agents understand a website’s most important content. |
| WebMCP | Browser-side pattern for exposing website functionality as structured tools rather than forcing agents to rely only on scraping or DOM navigation. |
| Agent Skills | Lightweight format for packaging task-specific instructions, scripts, templates, and resources that agents can use across contexts. |
| OAuth 2.0 Protected Resource Metadata | Standard for publishing metadata about OAuth-protected resources. Relevant for agent authentication and protected API discovery. |
| Agent-to-User Interface Protocol | Protocol for agents returning structured user interfaces rather than only text. Relevant for agent-native product experiences. |
| Agent-User Interaction Protocol | Event-based protocol for streaming agent state, text, tool calls, and UI interactions to frontends. |
| Universal Commerce Protocol | Open standard from Google and Shopify for agentic commerce — discovery, checkout, orders, identity. Built on REST/JSON-RPC with MCP, A2A, and AP2 support built in. |
| Payments and commerce protocols | |
| Machine Payments Protocol | Open protocol for machine-to-machine payments. Designed for APIs, tools, and content that agents can pay for directly in the request flow. |
| MPP specifications | Technical specification surface for MPP, including payment authentication, charge, discovery, MCP transport, and payment-method-specific flows. |
| x402 | Internet-native payment protocol based on HTTP 402 Payment Required. Important for paid APIs, content, services, and agent-to-server transactions. |
| Agentic Commerce Protocol | Open commerce protocol for agent-mediated shopping and checkout while preserving merchant control over products, orders, and payment flows. |
| Stripe Agentic Commerce | Stripe’s canonical documentation for enabling secure commerce between buyers, agents, and businesses. |
| Agent Payments Protocol | Protocol for agent payments with authorization, mandates, and auditability. Relevant where agents transact on behalf of humans or businesses. |
| AI platform and developer resources | |
| OpenAI Developers | Official developer hub for OpenAI’s agent-facing surfaces, including apps, tools, commerce, and ChatGPT-connected product patterns. |
| OpenAI Apps SDK | Canonical source for building ChatGPT-connected apps backed by MCP servers and machine-native interfaces. |
| Google Developers Blog | Official source for Google’s agent protocol, interoperability, and developer platform updates. |
| Google Agent Protocols Guide | Useful overview of how MCP, A2A, UCP, AP2, A2UI, and AG-UI fit into agent architecture. |
| Anthropic Engineering | One of the highest-signal primary sources on agent design, tool design, context engineering, MCP, skills, and autonomous coding systems. |
| Cloudflare Agents | Cloudflare’s developer platform for building and deploying agents, including Workers, Durable Objects, tools, scheduling, and MCP support. |
| Playwright MCP | Browser automation through MCP. Relevant for how agents interact with websites through structured browser control. |
| Vercel agent-browser | Browser automation CLI for agents. Useful reference for token-efficient browser interaction and headless browsing workflows. |
| Discovery, readiness, and machine-readable web | |
| Markdown for Agents | Cloudflare documentation for serving Markdown versions of pages to agents through content negotiation. |
| Content Signals Policy | Policy mechanism for expressing preferences around search, AI input, and AI training usage. |
| Web Bot Auth | Proposal for cryptographically identifying legitimate bots and agents rather than relying only on user-agent strings or IP lists. |
| Schema.org | Shared vocabulary for structured data. Important for making products, organizations, people, events, and actions machine-readable. |
| JSON-LD | Linked-data JSON format commonly used with Schema.org. Useful for giving agents explicit semantic context on web pages. |
| Hybrid companies at the edge of the headless economy | |
| Stripe Blog | Primary source for Stripe’s thinking on agentic commerce, payment infrastructure, internet-native business models, and merchant readiness. |
| Ramp for Agents | Strong example of incumbent business software becoming explicitly agent-facing through CLI and MCP-style surfaces. |
| Ramp CLI | Concrete implementation of a finance product exposed through a terminal and AI-agent-usable command surface. |
| Ramp Blog | Useful source for Ramp’s agentic finance, AP automation, accounting agents, and operational AI product direction. |
| Cloudflare Blog | Infrastructure source for how the web adapts to agents: MCP, bot identity, content negotiation, Workers, and agent deployment. |
| Linear | Product company building toward human-and-agent workflows in software development and product operations. |
| Linear MCP Docs | Direct source for Linear’s MCP server and agent-usable product surface. |
| Notion | Workspace product moving toward AI agents, automation, enterprise search, and machine-mediated work. |
| Notion MCP Docs | Official documentation for Notion’s hosted MCP server, allowing AI tools to securely access Notion workspaces. |
| Notion Blog | Ongoing source for Notion’s AI workspace, custom agents, and agent-native product direction. |
| Slack Blog | Official source for Slack’s evolution into an AI context layer and agentic work surface. |
| Slack MCP Guide | Practical source for how Slack exposes workspace context and actions to AI assistants through MCP. |
| Related blogs and analysis streams | |
| Fintech Brainfood | Simon Taylor’s recurring analysis on fintech, payments, AI, stablecoins, wallets, and infrastructure shifts relevant to agentic commerce. |
| MPP Blog | Ongoing protocol source for machine payments, per-request pricing, paid APIs, and agent-to-service payment flows. |
| Simon Willison’s Weblog | High-signal independent source on LLMs, agents, MCP, tool use, prompt injection, and machine-consumable software patterns. |
| Latent Space | Technical newsletter and podcast covering agents, AI engineering, infrastructure, MCP, and developer ecosystem shifts. |
| Every | Recurring source on AI product design, agent-native software, and emerging business models around AI workflows. |
| Benedict Evans | Strategy source for interface shifts, platform transitions, value capture, and market structure in technology. |