OSTARA.WORK

Klasker Academy

Learn how to be machine-readable, agent-readable and trusted by AI.

Artificial Intelligence is changing how people discover information, products and services. Therefore, AI agents increasingly need more than webpages: they need structured information, reliable interfaces, trustworthy identities and secure ways to act, in a modular fashion.

The Agentic Web

Originally, the web was designed for people. The emerging Agentic Web must also communicate effectively with software capable of discovering information, making intelligent decisions and personalised choices based on your own lifestyle and carrying out tasks on your behalf.


Agentic AI is the next generation of AI, as it learns iteratively, refining its approach to better align with its defined objectives. Agentic AI reasons on the steps to take and the results of its actions, plans and acts on tasks in response to a question or prompt (not only from humans). Agentic AI uses tools and is able to collaborate with other agents, and it can be so proficient on a certain domain that it can act an integral part of a virtual workforce, in partnership with human workers.


We are entering an era where AI agents will not just assist you — they will decide for you. Business models need to evolve from optimising clicks (on a website) to earning trust from algorithms acting for consumers. How do you make your product or service ‘agent discoverable’? How do you build a business in a new world where the buyer is a AI model acting in someone’s best interest? What will happen to brand loyalty? Do you have to create a ‘customer experience‘ for people or their AI agents? Like it or not, as time goes by, the AI agent will be your future ‘new customer‘.

Discover

How do AI agents find websites, products, services and business information?

Understand

How can machines reliably understand products, prices, availability, policies and business information?

Trust

How can an agent determine whether a business, domain, data source or transaction can be trusted?

Act

How can an AI agent move from reading information to interacting with a website and completing a task?

Learn

Explore the technologies and concepts behind Agentic AI and machine-facing digital trade.

Foundations

Understand Agentic AI, AI agents and the transition from conventional websites towards machine-oriented digital commerce.

  • What is Agentic AI?
  • What is an AI agent?
  • The Agentic Web
  • From search to action
Explore Foundations

Agentic Commerce Protocols

Learn how emerging protocols allow businesses, AI agents and payment networks to communicate and transact.

  • ACP
  • UCP
  • AP2
  • MPP
  • Visa TAP
  • Mastercard Agent Pay
  • MCP
Explore Protocols

Machine Legibility

Learn how websites can make their information understandable and useful to AI systems without relying exclusively on visual presentation or JavaScript rendering.

  • Schema.org
  • JSON-LD
  • Product data
  • APIs and feeds
  • llms.txt
  • Promise Drift
Explore Machine Legibility

Agent Trust

Explore the signals that allow digital businesses and their infrastructure to establish identity, security and trust.

  • Digital identity
  • DNS and DNSSEC
  • SPF, DKIM and DMARC
  • Authentication
  • Agent identity
  • Payment trust
Explore Agent Trust

Agent Readiness

Discover the difference between a website that appears machine-readable and one that an actual AI agent can successfully use.

  • Static analysis
  • Agent interaction
  • Task completion
  • Commerce workflows
  • Failure points
Explore Agent Readiness

Klasker Methodology

Understand how Klasker evaluates websites, products and services for the emerging environment of Agentic AI.

  • Protocol Coverage
  • Machine Legibility
  • Trust
  • Commerce Readiness
  • Agent Usability
Explore Methodology

Agentic Commerce Protocol Stack

Agentic commerce is not based on a single protocol. Different technologies address discovery, data access, commerce, authentication and payment.

ACP

Agentic Commerce Protocol — a protocol for enabling AI agents to interact with commerce systems.

UCP

Universal Commerce Protocol — an emerging framework spanning discovery, commerce and post-purchase experiences.

AP2

A payment-oriented layer designed to address secure, agent-authorised transactions.

MPP

Machine Payments Protocol — an emerging approach to payments between machines and services.

Agent Identity

Emerging payment-network mechanisms such as Visa Trusted Agent Protocol and Mastercard Agent Pay address agent authentication and trusted transactions.

MCP

Model Context Protocol — a standardised way for AI systems to access structured information and capabilities.


Klasker can use protocol detection as part of its broader evaluation of machine-facing digital readiness.

Machine Legibility

A website may look excellent to a person while still being difficult to understand for an AI agent.

Structured Data

Product, Offer, Organisation, AggregateRating and Availability information can provide machines with explicit meaning.

APIs & Feeds

Clean machine-facing interfaces can provide more reliable access to information than scraping dynamically rendered pages.

Answer Engine Optimisation

Learn how machine-readable information changes the way businesses should think about discoverability in answer engines and AI systems.

Promise Drift

When machine-readable information disagrees with live website information, agents may receive conflicting signals about price, availability or other commercial facts.

Can an AI Agent trust Your Business?

Agentic commerce requires more than useful information. Businesses must also establish reliable digital identity, infrastructure and transaction trust.

Domain & Identity

DNS, certificates, DNSSEC and consistent business identity form part of the digital environment surrounding a business.

Email Infrastructure

SPF, DKIM and DMARC provide important signals about the authenticity and handling of domain email.

Authentication

Agentic systems introduce new questions around authentication, authorisation and the identity of automated actors.

Transaction Trust

Payment and commerce protocols increasingly need mechanisms that allow agents to act within clearly defined authorisation boundaries.

Static Readiness vs Agent Readiness

A checklist can establish that the required signals exist. An actual agent interaction can reveal whether those signals translate into a usable experience.

Static Evaluation

Inspect HTML, structured data, headers, APIs, feeds, .well-known resources and protocol indicators.

Agent Evaluation

Give an agent a task and observe whether it can discover, understand and complete that task successfully.

Task Success

Measure whether the intended outcome was achieved and identify the steps at which an agent failed or became uncertain.

Commerce Workflow

Future Klasker evaluations can examine product discovery, carts, shipping, checkout and other transaction workflows.

How does Klasker evaluate a Website?

Klasker AITA combines observable technical signals with practical analysis of how digital information can support Agentic AI.

Protocol Coverage

Which relevant agentic protocols and machine-facing interfaces can be detected?

Machine Legibility

Can an AI system reliably understand the information presented by the website?

Trust

What technical and identity signals contribute to confidence in the business and its digital infrastructure?

Commerce Readiness

Can machine-facing information support a commercial interaction from discovery towards transaction?

Agent Usability

Can an AI agent successfully accomplish useful tasks?

From Academy to Observatory

Academy AITA explains the technologies and methodology, but the Observatory applies those ideas to real websites, products and services.

Explore public evaluations and observe how Agentic AI readiness develops across different industries and businesses.

Explore Observatory

Put the Knowledge Into Practice

Learn what matters, then evaluate how well a real website, product or service performs against the emerging requirements of Agentic AI.

Free Public Evaluation View Observatory

Academy AITA

Academy AITA is your learning hub for Artificial Intelligence Trade Analysis, combining public research with practical evaluation of the digital environment.