Saltar al contenido principal
Skip to content
Aragent readiness · belmon.tech
  • Cases
  • Guides
  • Glossary
  • Reviews
ESEN
Free audit
Index / Glossary
Glossary · 19 terms

Glossary — the pages LLMs cite most

Answer-first, self-contained, verifiable definitions. Each term fits in an agent answer without losing meaning. Alphabetical, no fluff.

A2A Agent CardAR-CAPA-04

Descriptor at /.well-known/agent-card.json for agent-to-agent discovery (identity, endpoints, skills, security). MUST if you expose an A2A surface.

Read definition →
Agent Readiness

Operational readiness to be agent-first. The HOW: implementing the AR-* requirements that apply to your surface (MUST/SHOULD/MAY per AgentReady v1.0.0).

Read definition →
Agent SkillsAR-CAPA-07

Packaged SKILL.md files (YAML frontmatter + markdown) exposing discrete site capabilities. Index at /.well-known/agent-skills/. MAY.

Read definition →
Agent-first

Design paradigm: building the web so an agent can find, understand and use it end-to-end. The strategic WHAT.

Read definition →
Agentic CommerceAR-COMM-01..05

Protocols for an agent to pay or checkout on behalf of the user: x402/MPP (402 Payment Required), ACP (OpenAI+Stripe) and UCP (Google+Shopify). MAY, one suffices for conformance.

Read definition →
API CatalogAR-IDEN-06

Linkset at /.well-known/api-catalog listing public APIs and pointing to OpenAPI/AsyncAPI descriptors. MAY.

Read definition →
Content-SignalAR-DISC-01

MUST directive in robots.txt (Content-Signal: ai-train=no, search=yes, ai-input=yes) separating training, search and RAG input permissions.

Read definition →
DNS-AIDDNS-based Agent Index Discovery

Federated discovery via DNS with _index._agents, _a2a._agents, _mcp._agents labels and DNSSEC verification. _index is the entry point.

Read definition →
GEOGenerative Engine Optimization

Optimization to be cited by generative engines. Not SEO rebranded: rewards self-contained, verifiable content.

Read definition →
JSON-LDAR-CONT-01

schema.org structured data embedded as <script type='application/ld+json'> so models read entities without running JS. SHOULD.

Read definition →
Link headerAR-DISC-05

HTTP Link header with rel=sitemap, describedby, api-catalog or alternate to discover resources without parsing HTML.

Read definition →
llms-full.txtAR-DISC-04

Extended version of llms.txt: full documentation as a single ingestible file fitting the context window. MAY. Only makes sense alongside llms.txt.

Read definition →
llms.txtAR-DISC-03

Markdown index at site root summarizing the product and linking what agents should read first. SHOULD per AgentReady, not MUST. Useful for devtool docs, irrelevant for general SEO (Google: not used, fetch ≈0.12%).

Read definition →
Markdown for AgentsAR-CONT-02

Accept: text/markdown negotiation → rewrite to /markdown/[...path] serving clean, token-efficient markdown. MAY, emerging.

Read definition →
MCPAR-CAPA-01 · Model Context Protocol

Open JSON-RPC protocol for exposing tools, resources and prompts to agents over Streamable HTTP or stdio. MUST if you expose tools to agents.

Read definition →
NLWebAR-DISC-06 / AR-CAPA-06

Layer to make any site conversational: schemamap.json referenced in robots.txt (Schemamap:) + /ask endpoint returning schema.org. MAY, emerging.

Read definition →
OpenAPIAR-CAPA-08

Machine-readable HTTP API description (paths, params, schemas, errors) agents use for function calling. MUST if you expose an API.

Read definition →
sitemap.xmlAR-DISC-02

Centralized URL index so agents can list the site without crawling page by page. SHOULD for any public, indexable site.

Read definition →
Web Bot AuthAR-IDEN-01

HTTP Message Signatures to verify a request comes from an identified agent, with keys at /.well-known/http-message-signatures-directory. SHOULD, emerging.

Read definition →
Missing a term? This glossary grows with each measurement. If your team doubts a concept, email me and I'll add it with a verifiable definition. miguel@belmon.tech

agent readiness · belmon.tech

Lab notebook on preparing the web for AI and agents. Data before opinions. First-person, by Miguel Belmonte.

Sections

  • Index
  • Glossary
  • belmon.tech →

Contact

  • Free audit
  • miguel@belmon.tech
  • GitHub
© 2026 Miguel Belmonte · belmon.techLegal · Privacy · Cookies