Learn

← Track G: Agentic Advertising: Foundations
L3-L4

G4

Trust, identity, and discovery for agents

In this lesson you will
  • Explain how a buyer agent discovers which sales agents are authorized to sell a property, using adagents.json as the agentic counterpart of ads.txt.
  • Connect agent authorization to the supply-chain transparency you already know: the same questions of who is allowed to sell, and how deep the chain runs.
  • Describe where human approval, brand safety, and fraud control sit in an otherwise automated flow.

A buyer agent wants to spend on a recipe site. Before it sends a single offer it has to answer one question, the same question a careful buyer has always asked: who is actually allowed to sell this site’s inventory? Get that wrong and you are paying a stranger.

You met this exact question in Track C. A publisher declares its authorized sellers in ads.txt, and a buyer verifies them against each SSP’s sellers.json. The agentic world needs the same trust layer, one level up: a way to declare which agents may sell a property.

adagents.json: ads.txt for agents

A property owner publishes an adagents.json file that lists the sales agents authorized to represent its properties. A buyer agent reads it to discover who it can legitimately deal with, then queries those agents for availability. It is the agentic counterpart of ads.txt: a public, machine-readable declaration of "these parties may sell on my behalf."

QuestionHeader-bidding answerAgentic answer
Who may sell this inventory?ads.txt (authorized sellers)adagents.json (authorized sales agents)
Is that party who they claim?sellers.json reciprocityAgent identity and registry checks
How deep is the chain?schain nodes and hopsThe agents and systems in the path
How a buyer agent gets to a safe buy
  1. 1Discover: read the property’s adagents.json to find authorized sales agents.
  2. 2Verify: confirm each agent’s identity, ideally against a shared registry.
  3. 3Query: ask the trusted agents what is available and at what price.
  4. 4Approve: route anything that needs a human (budget, brand safety) for sign-off.
  5. 5Execute: place the buy through an authorized agent only.

Watch out

An agent offering to sell a property that does not list it in adagents.json is the agentic version of an unauthorized reseller in ads.txt. It is exactly the kind of path the Bidcliq audit flags today, and the reason discovery has to come before the deal.

Key idea

Automation does not remove the human, it relocates them. People still set budgets and guardrails and approve sensitive steps. Brand safety and fraud control move from manual review to rules the agent enforces, with a person on the exceptions.

Quick check

A buyer agent is offered premium inventory by an agent that does not appear in the property’s adagents.json. What should it do, and what header-bidding red flag is this like?

Key concepts
adagents.json (authorized sales agents)Agent discovery and registriesAuthorization as the agentic version of ads.txtHuman-in-the-loop, brand safety, and fraud control
Going deeper

To really lock this in: given a property that lists two sales agents in adagents.json, explain how a buyer agent decides which to trust, and what would make an unlisted agent a red flag.

Where this leads
ANALYZE

The same trust question as the live audit: who is authorized to sell this inventory. Run an audit and read the authorized-seller checks, then picture the agentic equivalent.

Open →
PRACTICE

Proposed lab: author an adagents.json that authorizes a sales agent for a property.

Bidcliq Academy · G4: Trust, identity, and discovery for agents