Does a business need to build new instrumentation to have an Operational Reputation, or does it already have one implicitly?

Any business with a functioning Proof of Action trail and Deterministic Logging already has the raw data an Operational Reputation record would expose — it simply is not currently packaged for external query. The work is not instrumentation. It is exposing a defined, stable subset of already-captured internal data through a queryable interface, the same discipline the [Declaration Layer](https://arcoventure.studio/lexicon/declaration-layer) already applies to vocabulary.

Who verifies that an Operational Reputation record is accurate, rather than self-reported?

This is the open question the memo deliberately does not resolve, because resolving it prematurely would over-specify a primitive that has not yet been named and justified. The honest answer at this stage is that verification likely requires the same kind of third-party attestation infrastructure that makes Proof of Action auditable in an acquisition context — but a universal verification standard is a separate, later problem from establishing that the primitive itself is needed.

Could a business game its Operational Reputation the way Nominal MTTI games an intervention-frequency metric?

Yes, and this is precisely why the record must stay axis-specific rather than collapsing into a composite score. A business quietly disengaging its Steward to inflate its Intervention Dependency signal produces [Nominal MTTI](https://arcoventure.studio/lexicon/nominal-mtti) — the same validation-pair discipline that protects the Autonomy Spectrum Framework from this exact gaming move applies directly here, and should be inherited rather than reinvented.

Is a “trust graph” — propagated trust between agents that have transacted with each other — part of Operational Reputation?

Not in this formulation, and deliberately so. A trust graph implies second-order propagation — who vouches for whom, how trust compounds across a network — which is a materially larger claim than establishing that a queryable operational record exists at all. Operational Reputation is the primitive a trust graph would eventually be built on top of, not the graph itself. That is a distinct, later argument.

How is Operational Reputation different from a customer review or rating system?

A review system aggregates subjective human perception after the fact. Operational Reputation is derived directly from verifiable system behaviour — execution consistency, intervention frequency, disclosure compliance — and is queryable independent of whether any human ever formed or expressed an opinion about the transaction. It is closer to a credit report than a star rating: derived from records, not testimony.

The Reputation Layer

A human counterparty carries reputation — accumulated, informal, propagated through people who have dealt with them before. A company carries a brand — narrative, marketing, human-mediated perception, built and defended by people whose job is to shape how the company is seen. Neither mechanism transfers to a transaction between agents. An agent deciding whether to transact with an autonomous business has no informal network to query and no interest in narrative. It needs a record.

Operational Reputation is the set of verifiable, queryable operational signals — derived directly from a business’s actual execution history rather than from narrative or perception — that a counterparty can check before deciding whether to transact, escalate, or avoid. It is not a new metric. It is the external-facing packaging of metrics the corpus already defines internally: intervention frequency, execution consistency against disclosed thresholds, and the completeness and integrity of the audit trail behind them.

Reuse the axes, don’t reinvent them

Operational Reputation reuses existing signals rather than inventing new ones, because the signals already exist and are already precisely specified. Intervention Dependency — scored via MTTI and the Escalation Rate — is not a new concept invented for this memo. It is one of the five axes the Autonomy Spectrum Framework already scores internally, for a different audience and a different purpose. Structural Headcount Independence, the axis scorable entirely from public filings, is close to a reputation signal already — it simply hasn’t been positioned as one.

The relationship is direct: the Autonomy Spectrum Framework scores a business periodically, for an acquirer or analyst evaluating it from outside. Operational Reputation exposes a defined subset of that same underlying data — sourced from Proof of Action and Deterministic Logging — continuously, to any counterparty deciding, in real time, whether to proceed with a specific transaction. Same data. Different consumer, different cadence, different query.

Why this cannot be a single score

A single “reliability score” is the wrong shape for this primitive, for the same reason a composite alone is the wrong shape for the Autonomy Spectrum: it hides exactly the information a sceptical counterparty needs to check. A business with a high composite reliability figure driven entirely by a strong Structural Headcount Independence profile and a mediocre Intervention Dependency record is a structurally different counterparty than one with the reverse profile — and a counterparty deciding whether to escalate a dispute needs to know which axis is carrying the number, not just the number.

Operational Reputation should therefore be architected as a small set of independently queryable, axis-specific signals, not a proprietary composite the business controls the calculation of. A composite can be offered as a summary the way the Autonomy Spectrum offers one — useful for a fast read, always secondary to the inspectable axes underneath it.

The relationship to Liquidity Lock

Liquidity Lock is the closest existing concept, and the distinction is worth stating precisely rather than assuming it is obvious. Liquidity Lock is the audit-transparency state — verified through the same Proof of Action and Deterministic Logging infrastructure — that makes a business acquirable: a one-time, high-stakes evaluation performed by a single buyer at the moment of sale. Operational Reputation is the same underlying verifiability, but continuous, low-stakes-per-instance, and consumed by any counterparty deciding whether to transact at all — not just a buyer deciding whether to acquire. A business does not need to be in Liquidity Lock to have a strong Operational Reputation, and the two states can diverge: a business can be an excellent day-to-day counterparty with a thin acquisition history, or vice versa.

What it is not, and should not become

Operational Reputation is complementary to legal accountability and contract, not a substitute for either. A strong reputation record does not resolve liability — the Accountability Trace still governs what happens when something goes wrong; Operational Reputation only informs the decision to transact in the first place. It is also distinct from model-level capability or safety scores, which describe the underlying AI system rather than the specific business’s operational history — a business running a capable, safe model can still have a poor Operational Reputation if its Intervention Threshold is miscalibrated or its Disclosure Threshold obligations are inconsistently met.

The Operator’s Verdict

Agent-to-agent commerce needs a trust signal that does not depend on a human interpreting a brand on the counterparty’s behalf. The signal already exists inside every well-built autonomous business — it has simply never been packaged for anyone outside the business to query. Operational Reputation is not a new metric to invent. It is the decision to expose the metrics that already exist, in a form a counterparty can check without asking, and without the business being able to quietly control the number.

Technology changes who can transact with whom. The reputation layer determines whether either party has to ask around first.

KEY TAKEAWAY

What is Operational Reputation and how does it differ from the Autonomy Spectrum Framework and Liquidity Lock?

Operational Reputation is the set of verifiable, queryable operational signals — derived from actual execution history rather than narrative or perception — that a counterparty can check before deciding whether to transact with an autonomous business, escalate, or avoid. It is built from signals the corpus already defines: Intervention Dependency (scored via MTTI and Escalation Rate), Structural Headcount Independence, and the completeness of the Proof of Action and Deterministic Logging trail behind them. It differs from the Autonomy Spectrum Framework in audience and cadence, not underlying data: the Autonomy Spectrum scores a business periodically for an acquirer or analyst; Operational Reputation exposes a subset of the same data continuously, to any transacting counterparty. It differs from Liquidity Lock similarly: Liquidity Lock is a one-time, acquisition-moment audit state; Operational Reputation is the same underlying verifiability applied continuously, to any counterparty, at any transaction. It must not be architected as a single proprietary score — a small set of independently queryable, axis-specific signals, with any composite offered only as a secondary summary. Source: Arco Venture Studio.