Freeze the rules.
Every competition pins its engine, rules, generator, adapter version, policy, and evidence semantics. Historical runs do not silently inherit tomorrow's code.
LedgerLayer brings communities, resources, multiplayer and verifiable history into one place. Enter the Network—or keep your own server and explore a supported Verify integration.
EXPLANATORY EXAMPLE · Sample actions, hashes and receipt states illustrate the flow.
19e4a1+1JUMPb9517dseq: 217JUMPGames, lasting discussions, communities, resources and server history share one home.
Explore the Network →Verify follows the supported adapter and evidence contract. LedgerLayer hosting is an optional path.
Explore existing-server Verify →Every competition pins its engine, rules, generator, adapter version, policy, and evidence semantics. Historical runs do not silently inherit tomorrow's code.
The canonical transcript contains the initial state, ordered admitted actions, and terminal boundary — not a self-reported final score.
A verifier replays the transcript under the pinned environment and derives the result. Receipt claims say exactly what was and was not proven.
Portable evidence bundles can be independently checked outside the hosted platform. Open verification is part of the trust model, not an export checkbox.
This scripted example shows the evidence model. Change one sample action to see a mismatch propagate through transcript identity, replay, result binding and receipt validity. No remote verification job is submitted.
8d4a0c+1JUMP48b1feJUMPtarget91ae13+1tick 792 / JUMP
Every claim currently resolves against the canonical transcript and pinned environment.
Try it: one action changes. The proof notices.
Keep your game infrastructure. Submit completed runs. LedgerLayer independently replays the transcript, derives the result, signs the receipt, and produces portable evidence.
No gravity constants in the platform. No Snake body rules. No card effects. LedgerLayer handles identity, admission, sessions, evidence, receipts, leaderboards, and policy. Versioned adapters own gameplay semantics.
Fixed tick · procedural · ranked · ghosts
REFERENCE / REALTIMEGrid lockstep · deterministic · agent-ready
REFERENCE / AGENTLegacy engine · historical replay · migration
REFERENCE / LEGACYImplement the contract. Keep your engine.
THIRD-PARTYFor an already deterministic game, the integration goal is intentionally small: expose state creation, actions, stepping, result derivation, and manifests. LedgerLayer supplies the conformance harness around it.
Time to first verified run
The reference suite is deliberately heterogeneous. Each game exists to attack a different assumption in the platform — realtime, grid lockstep, legacy engines, compact turn-based transcripts, and eventually hidden information.
Each reference integration attacks a different assumption. The platform earns broader claims only when the same contract survives different runtime models, evidence strengths, and participant types.
Agents discover a versioned environment, receive legal observations, decide under deadlines, and leave behind the same replayable evidence chain as human competition.
The public verifier turns opaque leaderboard rows into inspectable claims. This sample shows how a receipt separates checked assertions from weaker claims. Inspect the actual evidence and method in Verify; the values below are illustrative.
Run 7K9P2M replayed to the signed terminal boundary under its pinned environment.
The commercial product is only credible when historical semantics, signing infrastructure, customer isolation, evidence retention, and recovery all survive real operational failure.
Registered engine identities resolve to fixed digests — never whichever binary happens to be current.
Third-party engines execute away from the public API under resource, filesystem, network, and runtime constraints.
Historical public keys remain available so old receipts can continue to verify after operational key changes.
Evidence identity comes from cryptographic digests, with independent mirrors and restore drills rather than filenames.
Storage, credentials, keys, usage, and operator access are scoped to explicit tenants unless an artifact is intentionally public.
Operator intervention becomes an auditable event. Competitive history is not silently rewritten in place.
The first commercial motion is intentionally high-touch. We are looking for teams with a deterministic or server-authoritative game, simulation, or agent benchmark where the result matters enough to preserve.
Build locally against the public contract before production infrastructure becomes part of the equation.
Work directly with LedgerLayer to connect one meaningful production workload and stress the platform outside our own reference games.
Ship the smallest complete trust loop first. The business does not need to wait for the entire five-year architecture before customers can verify real production runs.
Accounts, permissions, search, safe resources and clear proof scope.
One game, communities, useful files, actual play and lasting history.
Independent authors, community ownership and reliable optional hosting.
Locked rules, scoped results, private disputes and auditable corrections.
Explicit disclosure, independently checked claims and developer integrations.
Worlds, federation and independent services when their gates are met.
Start with one run. Pin the environment. Replay the transcript. Derive the result. Sign only the claims the evidence supports.