spark.exposed

Operators, hosted services, and refusal controls

Centralization

How the published Operator set, coordinator, hosted SSP, wallet gate, and adjacent compliance systems shape default-path control.

How to read this topic: 5 principal evidence groups organize 19 supporting records by the question they answer.

Critical boundary: The record documents defaults and capabilities. It does not show common ownership of Operator keys, coordinated enforcement, production use of the wallet gate, or a private compliance call wired into Spark.

Threat-model assumption TM-1: Lightspark–Flashnet coordination is treated as sufficiently likely to be a core adversarial assumption, creating one effective two-role quorum under the published threshold. Documented relationships make that risk material; they do not prove that collusion or shared key access has been observed.

Evidence group 01

Default coordination and service concentration

The reviewed SDK defaults fix Lightspark SO0 as coordinator without automatic failover and expose one Lightspark SSP, while allowing integrators to supply different configuration.

Scope boundary: Defaults shape the ordinary path but are not protocol-level exclusivity. A custom integrator can choose another coordinator or SSP.

Evidence group 02

Operator quorum under the Lightspark–Flashnet coordination assumption

The threat model treats Lightspark–Flashnet coordination as sufficiently likely to be a core adversarial assumption. The published two-of-three set then lets those roles form a threshold without Breez; in the cited SSP-funded receive scenario, matching retained historical keys would be sufficient for a conflicting spend.

Scope boundary: Coordination is a core adversarial assumption, not an observed event. The record documents working and personnel ties as context; the additional key-retention condition is also unobserved. No evidence shows a conflicting signature or loss.

Centralization

Retained Lightspark and Flashnet keys could sign a conflicting spend of a Lightspark-funded Lightning leaf

In Spark's reference inbound-Lightning flow, the SSP is the former owner of the leaf. Under the published two-of-three configuration, retained matching keys from the Lightspark SSP, Lightspark Operator, and Flashnet Operator would be sufficient to sign a fresh conflicting spend; no evidence shows retention or collusion. [9 sources]

Pinned source code · Official documentation · Blog or announcementRead the analysis

Evidence group 03

The wallet-specific refusal control

A per-wallet, per-Operator gate is implemented, tested, live-updatable under one provider, and enforced across several state-changing protocol paths.

Scope boundary: Each Operator controls its own gate. The public record does not show production activation, a network-wide value, its operator, or a sanctions-specific purpose.

Evidence group 04

The privileged SSP-to-Operator surface

The reviewed Operator policy defines a Lightspark-only, IP-restricted service surface with an application-anonymous subset for cross-system coordination.

Scope boundary: Application-anonymous does not mean publicly reachable. Method names and policy establish privileged access, not misuse or the live allowlist configuration.

Evidence group 05

Adjacent product and contractual context

Grid schemas, Grid product material, Connect screening code, and Flashnet terms document compliance and refusal capabilities around the Spark ecosystem.

Scope boundary: These records do not establish a private Grid, Connect, or Flashnet compliance decision wired into the Spark Operator wallet gate. They remain context, not proof of protocol enforcement.