Independent measurement for machine commerce

We measure what machine commerce actually does.

probe402 independently observes machine-payment infrastructure over time, retains the evidence, and publishes what the record shows — not what providers claim.

Read the latest research Request an audit

Explore the evidence record →

Why independent measurement

Discovery tells you what exists.
Providers tell you what they claim.
Payment records tell you what settled.
We measure what actually happened.

None of those says what happened when a machine actually tried to transact: whether the endpoint answered, what it quoted, whether the request could be built from what is published, and whether a payment bought anything usable. That gap is what probe402 measures, and the measurement is kept.

What we measure

Five questions, asked of known machine-payment endpoints on a clock. This is the instrument's field of observation.

  • Does it answer?
  • What does it quote?
  • Can an agent actually construct the request?
  • Does observed behaviour match the published claim?
  • Once a payment settles, is the answer usable?

What the record produces

One measurement capability feeds everything published here. Observations become retained evidence; the evidence supports published research and individual answers; and the same discipline is applied, on request, to a specific system as an independent audit.

Observation
The instrument records what each endpoint did at each check, from our own vantage, and never edits a reading afterwards.
Evidence
Every capture is kept and dated, and each closed archive day is sealed into a chain, so a finding can be checked against what actually came back.
Research
The Machine Commerce Reality Index and the published answers are written from that record, each tied to the dated evidence behind it.
Audit
The same measurement discipline, applied on request to one endpoint, host or question, as an independent audit.

Latest research

Machine Commerce Reality Index #001 ·

What machine commerce claims — and what actually happens

MCRI measures what publicly listed machine-payment infrastructure claims, what it actually does, and where the two diverge.

What the issue found

  • The observability layer is stale and wrong
  • Listed does not mean machine-executable
  • Infrastructure is more concentrated than catalogue diversity suggests

Read the issue →

Machine Commerce Conditions MCC-001 What Does a Machine Need to Know Before It Spends Money?

The evidence

Every finding has a record behind it.

Every published finding rests on an observation we made and kept. Endpoints are observed repeatedly rather than once; the bytes that came back are retained; every record is dated; the rule that graded it is versioned and published; a mistake is corrected in the open rather than rewritten; and each closed archive day is sealed so that it cannot be altered quietly.

Observed endpoints14,499on the seed list, each observed at least once
Sealed archive40closed archive days, sealed against quiet rewriting
Published corrections3filed against our own record, never rewritten

See how the evidence is kept → · Explore the evidence record →

Independent audit

Need to know what your machine-commerce infrastructure actually does?

Independent audits apply the same measurement approach to a specific endpoint, host or question, with additional scoped testing where needed.

Request an audit