kompiut_commercial_launch_readiness_v1

Commercial launch gate

Reading the live readiness contract.

kompiut_launch_readiness_bridge_v1

Public launch gate

A support-safe summary for humans and agents. It shows what is verified, what is blocked, and which no-write work can continue.

Overall readiness38%

blocked before public launch

ExecutionDisabled

Network writeNo request

DeviceNo device touch

ShareableSupport-safe summary

Trust, safety and redaction foundationverified foundation

70%

Run no-write launch-readiness bridge and verify every blocked live boundary.

Proof, verifier and runner loopverified foundation

65%

Prepare the exact first-proof-run launch bridge without touching the device.

Real device end-to-end executionblocked external

25%

Wait for exact target-device approval, then run one scoped proof exercise only.

Human and agent contributor UXpartial

35%

Add guided launch states that explain approval files, proof state and blocked live actions.

Billing, legal and payoutsblocked external

15%

Create a launch-blocking billing/legal checklist before any payout implementation.

Marketplace demand and pricingpartial

20%

Define one customer pilot path with allowed data, verifier, support owner and margin model.

Ops, support and scalepartial

35%

Turn launch readiness into a repeatable pass/fail operator drill.

kompiut_customer_pilot_launch_gate_v1

Customer pilot gate

Pilot evidence must name the data class, verifier, margin, support owner, incident path, legal state and redacted proof before customer claims widen.

Customer pilot0%

blocked before customer pilot

Customer computeNo customer compute

BillingDisabled

PayoutsDisabled

Pilot evidenceRequired

Next safe work

Operator launch-readiness bridge for the first real proof run
Human and agent guided contributor activation
Customer path launch gate

Blocked live boundaries

Real job creation: blocked
Target-device runner cycle: blocked
Runner output submission: blocked
Verifier worker operation: blocked

kompiut_customer_pilot_evidence_chain_bridge_v1

Evidence-chain bridge

Operator view for the private pilot chain. It keeps customer compute, billing, payouts, network writes and device touch disabled while showing missing proof layers. Pilot economics stay blocked until the trusted chain is attached.

Evidence chain0%

blocked missing trusted chain

Customer computeNo customer compute

BillingDisabled

PayoutsDisabled

Network writeNo request

Customer pilot guidance

Collect the missing trusted evidence in order before any customer-pilot claim widens.

Guidanceblocked

currentPilot path evidenceMissing one reviewed path; no live customer action can use this layer yet.
lockedSupport drill evidenceMissing same path; no live customer action can use this layer yet.
lockedBilling, legal and payout-readiness evidenceMissing same pilot/support chain; no live customer action can use this layer yet.
lockedRepeat demand and margin evidenceMissing same pilot/support/commercial chain; no live customer action can use this layer yet.
lockedPrivate pilot reviewReview can start only after path, support, commercial and demand evidence are all attached.

Missing evidence

No reviewed customer pilot path is attached.
No allowed customer data class is attached.
No verifier rule is attached.
No positive price and margin model is attached.
No support owner is attached.
No incident path is attached.
No trusted support drill evidence is attached.
No trusted billing, legal and payout readiness evidence is attached.
No trusted repeat customer demand and unit-margin evidence is attached.
Legal and billing status are not approved for a private pilot.
No redacted proof evidence is attached.

Disabled pilot boundaries

Do not create live customer jobs from this bridge.
Do not enable customer compute, billing or payouts from this bridge.
Do not publish broad marketplace claims from this bridge.
Do not expose private path ids, aliases, margins, ticket hashes or proof hashes in public output.

Safe next action: Prepare pilot path evidence with the required bindings; keep it review-only until the admin dry-run passes.

Check my device

Bring a Mac mini, GPU box or server. Start with fit.

Kompiut first helps you understand what your device can safely contribute, what runner target it fits, what is missing and what stays out. Setup only comes after a contribution path has a reviewed trust model.

No hidden remote shell · No private keys · No personal files · No automatic mode in V1

01Choose a device

Start with Mac mini, Linux GPU box, gaming PC or server. Exact specs can wait.

02Add rough facts

Pick GPU, capacity, availability and installed tools with buttons.

03Check readiness

See blocked, fixable and unknown states before any setup.

04Choose a path manually

No automatic mode in V1. You decide what to run when a path is supported.

Contributor flow

Check device

active

Pick the closest device facts and run readiness.

01Check devicePick the closest device facts and run readiness.current
02Save profileReadiness must be checked before saving the device profile.locked
03Choose pathA saved device profile is required before path opt-in.locked
04Save opt-inChoose a matched path before saving path consent.locked
05Runner setupNo runner setup is available before saved path opt-in.locked

Proof approval guide

Start with device readiness. Later proof depends on saved facts and consent.

01Device factsLocal readiness facts stay visible before any path setup.current
02Path consentPath consent stays locked until a matched path is selected.locked
03One-shot approvalA separate approval must name the saved device, path, limits and expiry.locked
04Verified workOutput only counts after receipt and verifier evidence match the approval.locked

Runner setup approvalSetup approval stays locked until path consent is saved.

locked

Saved device
Selected path
Saved consent

Activation credentialActivation stays locked until setup approval is consumed.

locked

Consumed setup approval
Device-bound activation

Heartbeat approvalHeartbeat stays locked until activation is reviewed.

locked

Activated runner
Current consent
Stop/revoke state

Next-job approvalNext job stays locked until runtime readiness and heartbeat approval match.

locked

Runtime readiness
Consumed heartbeat approval
Scoped job

Output verificationOutput stays locked until receipt and verifier evidence match the approval.

locked

Output receipt
Verifier evidence
Matching approval

Guided activation

Complete local readiness and profile save before activation review can begin.

Proof stateNo proof state exists before saved consent and reviewed approvals.locked

01Device profile evidenceReadiness and saved profile facts must match the workspace.current
02Path opt-in recordPath opt-in stays locked until a reviewed path is selected.locked
03Setup approval fileSetup approval stays locked until path consent is saved.locked
04Activation proof stateActivation stays locked until setup approval and credential approval are both reviewed.locked
05Runtime job approvalHeartbeat and next-job approvals stay locked until activation proof is current.locked
06Verified outputWork only counts after output receipt and verifier evidence match the approval chain.locked

Approval files

Saved path opt-in readback
Runner enrollment setup approval artifact
Activation credential approval artifact
Runtime heartbeat and next-job approval artifacts
Output receipt plus verifier evidence

Blocked live actions

Runner enrollment
Activation credential issuance
Job claim
Output submission
Verifier worker
Billing and payout enablement

Finish device readiness, workspace access and profile save first.

Path opt-in

Choose a path before setup review.

Private beta opt-in is a separate consent step. Runner enrollment and setup still need their own reviewed approval before anything can run.

DeviceNot saved

WorkspaceChecking

PathNot selected

Only reviewed runtime limitsThe runner can only receive reviewed path jobs with limits.Declared network access onlyThis beta path uses the network rules shown above.No personal data or secretsPersonal files, cookies, keys and seed phrases stay out.Stop and uninstall stay availableYou can pause the path and remove the runner.No reward promise yetLedger is proof-only until payouts are enabled.

Preview only

No-write opt-in preview

This is the consent record review before save. It does not create runner setup, jobs, customer compute, billing or payouts.

WorkspaceChecking

DeviceNot saved

Reviewed pathSelect reviewed match

ConsentWaiting for confirmations

Runtime planSelect reviewed match

Data boundarySelect reviewed match

Network accessSelect reviewed match

Proof artifactSelect reviewed match

previewOnly=true
runnerEnrollmentCreated=false
setupCodeIssued=false
jobCreated=false
customerComputeEnabled=false
billingEnabled=false
payoutsEnabled=false

Device profileSave profile first

WorkspaceConnect access

Matched pathSelect reviewed match

Safety confirmationsReview every limit

Select a matched path first.

Commercial launch gate

Public launch gate

Customer pilot gate

Next safe work

Blocked live boundaries

Evidence-chain bridge

Customer pilot guidance

Missing evidence

Disabled pilot boundaries

Bring a Mac mini, GPU box or server. Start with fit.

Local device check

What do you have?

What kind of GPU?

How much room does it have?

When can it run?

What is already installed?