EXIT/ ROADMAP

PUBLIC DEVELOPMENT MAP · UPDATED 28 OCT 2026

Evidence becomes
infrastructure.

We are building an evidence-based blockchain research and monitoring system. Control and change analysis is the first specialization. Connected, verifiable investigations are the direction.

One master roadmap: the operated first release is followed by measured, continuous development toward evidence-based blockchain research and monitoring infrastructure. This map separates delivered work, the live first service and future growth. Every expansion has an evidence basis, a dependency and a way to test whether it actually helps.

Conceptual workflow, not a benchmark or a claim of universal coverage.

YOU ARE HERE

The first service is open.
The next job is usefulness.

The free campaign runs 25 October–24 November 2026, ending at 12:38:20 UTC. Solana mint and Ethereum contract reads, durable research records, on-request change checks, saved comparisons and answer-to-source navigation are live in the same service. Bounded authenticated operator acceptance is recorded; this is not a claim of continuous monitoring or universal coverage.

Open the current research interface ↗
Delivered in declared scopeLive / current windowPlannedLong-horizon direction
01FOUNDATION
DELIVERED / BOUNDED SCOPERetained development: 10 Jul–25 Oct 2026

Make an answer inspectable.

A research core that connects control relationships, sources, two kinds of time and later corrections.

  • Authority and control relationships for declared Solana subjects, with evidence and account-structure limits.
  • Distinct on-chain validity and evidence-known time; historical answers and corrections remain distinguishable.
  • Reproducible exports, deterministic tests and adversarial software-consumer checks.
  • Bounded Ethereum control and change cases; cross-chain work has begun, with explicitly limited coverage.
How we build it · how we know it works

MechanismTyped facts and relationships, source identities, temporal queries and replayable evidence. This is the foundation for a broader research graph.

AcceptanceSource-bound query and replay tests; missing, stale and contradictory evidence controls. Technical acceptance does not prove independent human usefulness or commercial demand.

DependencyDelivered results apply to their declared subjects, source versions and evidence windows. They are not a universal blockchain investigation capability.

02FIRST SERVICE
LIVE / ACCEPTANCE FOLLOW-UP OPENFree campaign: 25 Oct–24 Nov 2026

Bring your own Solana token.

Use the web interface or API to inspect current authority designations for a supported Solana token mint, beyond the original recorded examples.

  • Paste a Solana mint address: inspect mint, freeze and supported Token-2022 authority fields.
  • Observe supported authority-account structures in a shared finalized read, with slot and acquisition time.
  • Keep evidence references and explicit limits. Changed or incomplete acquisition is not a completed answer.
  • No wallet, token purchase or email required. A private API access key and fair-use quotas protect the free service.
How we build it · how we know it works

MechanismA bounded read-only acquisition step feeds existing control interpreters. Recorded historical questions remain a separate mode; arbitrary wallet forensics and free-form investigations are not yet supported.

AcceptanceDeployment and public route checks passed. The new route's final authenticated operator acceptance is still open in the work list; a live feature flag alone does not close that check.

DependencySupported parsed SPL/Token-2022 mint accounts, provider availability and the published campaign limits. API Pass payment and credit redemption are not live.

03Q1 · MONTHS 1–3
CURRENT GROWTH WINDOW / PARTIAL DELIVERY25 Oct 2026–25 Jan 2027

Turn a query into continuing research.

A useful investigation should survive the first answer: save its context, inspect its evidence and return when new evidence arrives.

  • Live: trackable research records for evidence gaps, with bounded follow-up and a clear resolution or recorded reason. Next: measure and improve supported task completion.
  • Saved context, source navigation and two-snapshot comparisons with their limits intact.
  • Scoped watch lists and pull-based change views; no claim of continuous alerts before the operating profile is verified.
  • Documented tools for software/agent consumers; measure task completion, unsupported claims, freshness, latency and cost separately.
How we build it · how we know it works

MechanismReuse the query, temporal evidence and export core. Add typed question plans and permission-bounded persistence. Constrained language assistance may help plan a question; it cannot invent unobserved facts.

AcceptanceEnd-to-end user tasks with missing/stale/conflicting-source counterexamples; account isolation, revocation and recovery tests. Research records are live in the free service; automatic alerts run as a limited pilot (one watched subject service-wide, read hourly). Wider alerting remains planned until implemented and verified.

DependencyThe operating first service, safe schema migration and permitted retained content. A stronger free workflow is the priority; future breadth is not a gate delaying current usefulness.

04Q2 · MONTHS 4–6
PLANNED / DEPENDENCY-BOUND25 Jan–25 Apr 2027

Connect authority, events and time.

Move from isolated answers toward an inspectable temporal research graph across selected blockchain surfaces. A first bounded graph, joining up to four of an account's own answers, is already live in the free service; the broader graph below remains planned.

  • Join typed authority and event relationships without confusing addresses, networks or time windows.
  • Add one bounded network or control surface selected for data quality, permissions, user need and acquisition cost.
  • Make source disagreement, missing history, freshness and finality visible in graph paths.
  • Introduce fund-flow or bridge relationships only with their own verified semantics and explicit evidence.
How we build it · how we know it works

MechanismChain-specific adapters and exact entity identities feed a common evidence-backed relationship model. Cross-chain research is selective expansion, not an assertion of complete coverage of every chain.

AcceptanceGolden graph/path tasks; wrong-network and incorrect-join counterexamples; reorganization and conflicting-provider tests. An address relationship never establishes personal identity or complete net profit.

DependencyStable Q1 contracts, permitted historical data, measured costs and an accepted bounded scope.

05Q3 · MONTHS 7–9
PLANNED / MEASURED RESEARCH25 Apr–25 Jul 2027

Investigate market patterns responsibly.

A bounded Market Integrity module that helps researchers test possible explanations rather than treating every unusual pattern as manipulation.

  • Replayable transaction and liquidity timelines for declared event classes and observation windows.
  • Explicit alternative explanations for coordination and anomaly signals.
  • Quality reports that show false positives, missed events, unknown labels and drift.
  • Real-time claims only after a separate capture, freshness and latency profile passes.
How we build it · how we know it works

MechanismDeclared time-window features and held-out reference periods, built on the existing Market Integrity research direction. A research signal is a starting point for inquiry, not a verdict.

AcceptancePrecision and recall only for labelled reference sets with clear denominators and windows. Test outages, drift and non-adverse explanations; disclose where ground truth is absent.

DependencyValidated Q2 data, permitted representative history and an accepted acquisition-cost budget. No universal manipulation detection, crash prediction or identity attribution is promised.

06Q4 · MONTHS 10–12
PLANNED / ORGANIZATIONAL WORKFLOWS25 Jul–25 Oct 2027

Make research usable across teams.

Permission-scoped workspaces, integrations and reproducible reports for a selected organizational workflow.

  • Workspaces with least-privilege access and research continuity across authorized users.
  • Versioned APIs and report delivery that preserve source permissions and reproducibility.
  • Operational controls for account separation, revocation, restore and capacity.
  • Conditional R&D: at most one bounded prototype in formal authority reasoning or verifiable computation.
How we build it · how we know it works

MechanismBuild on stable service contracts and evidence records. A proof applies to its declared model or computation; it does not prove that every input is complete or every economic explanation is true.

AcceptanceCross-account isolation, recovery and equivalent permitted-answer tests; a concrete workflow must improve before it is called delivered.

DependencyStable service operation, demonstrated need, data rights and resources. Research prototypes are not promised production breakthroughs.

07BEYOND YEAR ONE
LONG-HORIZON DIRECTION / NO FIXED DEADLINEContinuous, evidence-led development

A verifiable research layer for blockchain activity.

A connected research environment where people, software and organizations can ask supported questions, inspect the evidence and understand what remains unresolved.

  • Broader validated cross-chain research graphs for authority, events and observed fund relationships.
  • Inspectable planning: a question becomes a bounded, testable research plan with known source and cost limits.
  • Supported scenario analysis and stronger machine-verifiable, replayable outputs.
  • A continuously improved service whose coverage and quality are disclosed rather than assumed.
How we build it · how we know it works

MechanismCombine the proven control/change core with selectively validated data, graph, planning, monitoring and organizational capabilities. Scale the evidence and reasoning, not unsupported certainty.

AcceptanceExpand only when the same declared task improves in correctness, completion, speed or cost without hiding regressions. Stop, narrow or reformulate increments that fail that test.

DependencyData permissions, source quality, real task evidence, resources and sustained operation. This is a North Star, not a final state where research or engineering ends.

THE BIG PICTURE

More connected evidence.
More useful questions.
Fewer unsupported conclusions.

01Evidence layerSource identity · history · freshness · conflicts
02Research layerControl · events · typed relationships · time
03Planning & monitoringBounded questions · measured signals · recorded gaps
04People, software & organizationsWeb research · APIs · workspaces · permitted reports

Conceptual destination. The layers above combine delivered foundations with planned capabilities; this is not the current service architecture or an implementation disclosure.

HOW THE WINDOWS WORK

A plan is a testable direction.

Q1–Q4 are successive three-month windows from the token milestone on 25 October 2026, not calendar-year quarters. They are resource-dependent planning targets, not guaranteed completion dates. An unfinished outcome stays unfinished when a window ends.

At each step, compare answer quality, unsupported claims, completion, latency and cost separately. Publish actual outcomes and explain moved targets. Human usefulness and commercial demand require their own evidence.

EXIT does not promise to identify every manipulator, infer the person behind every wallet, predict crashes or guarantee investment outcomes. A research signal is not an accusation.

READ THE DETAILS

Illustrated whitepaper ↗Public roadmap document ↗Recorded cases ↗Free research ↗

Based on the accepted growth program and the current service records, checked 27 October 2026. English edition. Status labels describe this dated publication; later releases may change them.