Product proof

# Proof should answer a buyer question—not advertise a test count.

Engineering verification is necessary but is not a customer outcome. Accordo publishes decision-relevant questions and links each answer to current evidence and its boundary.

Buyer questions

## What can the system establish independently?

Numeric benchmark results appear only after a protocol, environment, version, date, limitation and reproduction path exist.

### Can policy reject an invalid decision?

Commercial policy is deterministic code, not a model's judgement: a renewal at or above the threshold stops and waits for a named human.

Benchmark dimension: Invalid-decision refusal

Limit: Proven for the built-in renewal object and its single value threshold. A general policy engine over arbitrary custom objects does not exist.

[Inspect evidence](evidence.html#C-03)

### Can a human gate stay outside agent authority?

The agent cannot approve on the human's behalf. A test asserts the refusal, so the boundary is a property of the system rather than a promise in a README.

Benchmark dimension: Human authority boundary

Limit: In local-development mode the actor is asserted, not authenticated: no authentication ships, so an actor header there is not an identity. This holds a boundary against an honest agent, not against an attacker with network access.

[Inspect evidence](evidence.html#C-04)

### Can the application describe itself?

One command tells an agent what an application actually is — packages, capabilities, resources, actions, policies, providers — read from checked-in source, in a single deterministic JSON report.

Benchmark dimension: Independent composition inspection

Limit: Source-only and read-only. It never opens the database, contacts a provider, reads a secret, or reports runtime, CI or authorization state — and it lists those blind spots as machine-readable limitations in its own output.

[Inspect evidence](evidence.html#C-14)

### Can the result run without a hosted vendor runtime?

Zero third-party runtime dependencies. Node 22 and a checkout — no build step, no bundler, no framework underneath your framework.

Benchmark dimension: Source portability and hosted-runtime independence

Limit: Development dependencies and the eventual PostgreSQL adapter are separate questions. Having no runtime dependencies is a property of the framework, not of whatever you add on top of it.

[Inspect evidence](evidence.html#C-17)

Implementation truth stays separate. Current proof and limitations come from the generated [claims ledger](evidence.html), not from this product description.

Benchmark architecture

## Product, agent and engineering benchmarks stay distinct.

A synthetic run is never presented as customer evidence. Where no equivalent baseline exists, comparisons cover architecture, responsibility and verification rather than invented competitor numbers.

[Open the exact repository claims and limitations](evidence.html)

---
Canonical: https://accordo.dev/proof.html
Implementation evidence: https://accordo.dev/claims.json
Build provenance: https://accordo.dev/version.json