Causalea is the public doorway into KELLY.

KELLY is the governed reality-learning engine. Causalea organizes the public windows: ask questions, compare forecasts, inspect markets, and open the proof shelf.

Public users can inspect and ask. They do not directly write KELLY's ground truth.

Governed by CANONICAL-BUILD-LAW-v1|External intelligence claims removed.|All public views are read-only inspection surfaces.

Governed by CANONICAL-BUILD-LAW-v1. All public windows are read-only inspection surfaces.

Learning Loop Passport
checkingpublic no-write

Current public-safe loop

live_brain_memory_worker

waiting_for_reality

Checking the next public-safe loop snapshot.

memory receipts

0

replay-ready closures

0

rollback ready

0

quarantine ready

0

Next proof checkpoint

checking

Source

KELLY keeps moving through approved source intake, hypothesis creation, reality closure, replay receipts, and memory-cell receipts. Users can watch and ask; public users cannot write ground truth.

Worker statusFeed matrixResearch queue
1

Source

checking
2

Observation

checking
3

Hypothesis

checking
4

Proof

checking
5

Settlement

checking
6

Replay

checking
7

Memory

checking

Passport progress: 0/7 stages visible. This is proof visibility, not a public mutation control.

Open full Glass Lab

Public KELLY windows
checking

These are the public read-only windows into KELLY. They let users ask, forecast, inspect markets, and watch proof without directly writing KELLY's ground truth.

Contract: public-surface-proof-contract-v1 | Public write allowed: false

Checking live windows...

Last checked: not yet

Live lens contract
checking manifest

This is the public app map KELLY exposes to the website, static Railway surface, and live monitors. It shows which standalone labs are canonical read-only lenses and which proof endpoints make each one traceable.

Public lens readiness

checking

Checking the compact public readiness route across the live lens layer.

public_lens_live_contracts

Apps

...

... repos

Contract endpoints

...

www, apex, Railway

Live app contracts

...

standalone lens control APIs

Readiness APIs

...

live lab health contracts

Artifact spines

...

source-to-memory trace APIs

In-lab Ask lanes

...

sector answers stay in context

Cognition organs

...

science, hypothesis, memory

Public mutation

unknown

ground truth protected

Manifest: config/public-lens-apps-v1.json | Route: /api/public-lens-apps | Contract rollup: /api/public-lens-contracts | Readiness: /api/public-lens-readiness

Standalone user-control contracts

These rows prove which deployed lens apps expose export/delete/location/privacy boundaries right now.

Open rollup JSONOpen readiness JSON

User controls across every lens

Settings is the shared control room for the public viewport: export browser-local data, delete local user info, manage coarse private location, change appearance, and send feedback. These controls never write KELLY Prime memory, GraphState, providers, or canonical reality.

Open SettingsLegal and privacy boundaryOpen Source & Third-Party Notices

Ask KELLY

Ask in plain language. KELLY answers with a bounded response, risks, next step, and proof path when available. Users guide attention. They do not directly mutate KELLY memory, GraphState, or canonical reality.

Shows the answer chain without letting a public user write ground truth. Proof view: /glass-lab#prime-research-queue

Superforecasting

Compare a user forecast beside KELLY's independent forecast, then score the outcome when reality settles. A user forecast can become a search/probe request. It is not treated as reality.

Keeps user probability separate from KELLY probability so calibration stays honest. Proof view: /glass-lab#calibration

Markets Lab

Track default stocks and crypto with forecast cones, source links, proof links, and model paper P&L. Read-only analysis. Not financial advice, not trade execution, and not user-directed ground truth.

Every visible forecast should point back to source, timestamp, model posture, and settlement status. Proof view: /glass-lab#markets

Glass Lab

Watch KELLY ingest approved sources, form hypotheses, close loops, replay proof, and expose memory-cell visibility. Observer-only. Glass Lab shows what KELLY is doing; it does not let public users steer writes.

This is the live window into receipts, replay, event ledger, memory, calibration, authority, and link health. Proof view: /glass-lab#live-proof-index

Betting and Odds Lab

Inspect line movement, implied probability, KELLY posture, and settlement receipts. Read-only analysis. Not betting advice and not bet execution.

Every visible forecast points back to odds snapshot, probability conversion, and outcome receipt when available. Proof view: /glass-lab#odds

Weather Lab

Check hazard risk near you with proof-linked reads from NOAA, NWS, and public feeds. Read-only awareness. Not emergency instruction.

Every snapshot names source freshness, place authority, and the closure path back to Glass Lab. Proof view: /glass-lab#weather

Policy and Economics Lab

Inspect regulations, public spending, macro pressure, and policy uncertainty. Read-only research. Not legal, tax, or investment advice.

Every claim should tie to citations, effective dates, and source authority. Proof view: /glass-lab#policy

Science and Research Lab

Inspect papers, claims, contradictions, falsifiers, and research trail. Read-only research support. Not medical or scientific authority by itself.

Every research claim should expose source metadata and falsification posture. Proof view: /glass-lab#science

Hypothesis Lab

Watch observations become falsifiable hypotheses, source plans, and reality checks. Read-only inspection. Public users cannot approve memory or GraphState writes.

Every hypothesis should expose falsifier, source plan, settlement target, and replay path. Proof view: /glass-lab#hypothesis

Memory and Lineage Lab

Inspect memory cells, generations, parent-child links, quarantine, and proof ancestry. Read-only inspection. Memory writes are internal authority only.

Every cell should connect to its replay receipt, lineage, and quarantine posture. Proof view: /glass-lab#memory-lineage

Which window should I use?

I want to ask a plain question

Ask a real question, then save it as a watched question if the answer needs future evidence. Private watched questions stay in the browser. Only sanitized public-safe research intent can reach KELLY Prime.

Open Ask KELLY

I want to compare forecasts

Turn an uncertain future event into a forecast with a closing condition and settlement date. User probability and KELLY probability stay in separate lanes. User forecasts are never treated as ground truth.

Open Superforecasting

I want market predictions

Pick a tracked asset and inspect the source, forecast posture, proof link, and paper P&L. Market views are read-only analysis. User watchlists can guide attention but cannot write canonical reality.

Open Markets Lab

I want proof and transparency

Start with the Learning Loop Passport, then open the feed worker, Prime queue, and replay panels. Glass Lab exposes proof and status only. Public users cannot trigger writes from the console.

Open Glass Lab

The simple KELLY loop

  1. 1. Reality enters: Registry-approved source lanes and enabled live feed workers provide observations with provenance, timestamps, freshness, and authority tags.
  2. 2. KELLY forms questions: The engine generates hypotheses, scenarios, and forecastable claims from memory, sources, and causal structure.
  3. 3. Reality closes loops: Settlements and calibration compare claims against outcomes instead of treating confident text as truth.
  4. 4. Memory is governed: Only eligible closed loops can become durable memory cells through protected adapters and replayable proof.
  5. 5. Glass Lab exposes proof and uncertainty.
  6. 6. Calibration and governed memory preserve what was learned.

What happens after you ask?

  1. 1. KELLY answers from what it can currently support.
  2. 2. You can keep the question in a private browser watch lane.
  3. 3. KELLY Prime receives only a public-safe research intent.
  4. 4. Registry source lanes, enabled live feeds, source plans, and hypotheses appear in Glass Lab.
  5. 5. Reality settles the outcome when the answer sheet exists.
  6. 6. Replay, calibration, and governed memory decide what KELLY can retain.

Proof shortcuts

Live status

Fast view of worker health, public no-write posture, and the current learning loop passport.

Best first check when something feels stale.

Hypothesis Lab

Shows KELLY-generated hypotheses, Prime research intents, and memory birth candidates.

Best view for what KELLY is trying to learn next.

Feed Matrix

Shows which source families can support which question types.

Best view for source mode and intake coverage.

Data Worker

Shows accepted feed activity, carrying feeds, provider issues, and cooldowns.

Best view for whether reality fuel is flowing.

What users can ask KELLYplain language

KELLY is useful when the question can be bounded, checked, forecast, replayed, or settled later. Users do not need to know the internal machinery; the UI should show what KELLY did with the request.

Make a decision under uncertainty

Use KELLY when you need a next action, a risk check, and a reason to wait or move.

Should I act on this in the next 24 hours?
What outcome would make this decision unacceptable?
What should I test before committing?

KELLY does: KELLY frames the choice, names the uncertainty, checks available evidence, and returns a bounded answer.

Proof appears: Ask KELLY shows the answer; Glass Lab shows proof, replay, source mode, and protected boundaries.

Boundary: The user request guides attention. It does not directly write memory or GraphState.

Forecast a future event

Use KELLY when a question can later be scored against reality.

What is KELLY's probability that this event happens by Friday?
What would change the forecast most?
What should we watch before settlement?

KELLY does: KELLY keeps the user forecast and KELLY forecast separate, then prepares a settlement path.

Proof appears: Superforecasting shows probability lanes; calibration shows whether reality improved the model.

Boundary: A user forecast is a probe request. It is not treated as ground truth.

Inspect market predictions

Use Markets Lab for stock and crypto forecasts, source links, cones, and model paper P&L.

What is KELLY forecasting for this asset by end of day?
Which source moved this forecast?
How did KELLY perform on this ticker recently?

KELLY does: KELLY reads approved market data, separates quote from forecast, and tracks settlement/P&L.

Proof appears: Markets Lab should show source, timestamp, forecast, settlement state, and proof link per asset.

Boundary: Read-only analysis. Not financial advice and not trade execution.

Ask about data or sources

Use KELLY to turn approved data into observations, hypotheses, and reality checks.

What changed in this source since the last observation?
Which entities does KELLY recognize here?
What is stale, uncertain, or quarantined?

KELLY does: KELLY tags source mode, freshness, authority, uncertainty, and memory eligibility.

Proof appears: Glass Lab shows source-stage ledgers and whether a loop is waiting, closed, or memory-eligible.

Boundary: Public users cannot directly declare source truth. Approved data APIs and adapters carry authority.

Audit an answer

Use Glass Lab and receipts when you need to know why KELLY said something.

Show the proof receipt for this answer.
Can replay reconstruct the chain?
What did KELLY know, infer, and still not know?

KELLY does: KELLY exposes claim boundaries, proof receipts, replay status, settlement plan, and calibration state.

Proof appears: Glass Lab is the proof console; Replay rebuilds chains; Receipts show bounded evidence.

Boundary: Transparency is not a promise of certainty. Unknown or unscoreable claims must stay visible.

Private user lane

Users can have private questions, forecasts, and context beside KELLY. KELLY's shared learning comes from closed reality loops, not from exposing a user's private data.

Ask KELLY can be private to the user session or account.
User forecasts stay separate from KELLY forecasts.
A user question can guide KELLY toward a reality loop.
Shared KELLY memory should retain reality-closed lessons without leaking private user context.

What KELLY will not do from public UI

  • Public users cannot directly write memory, GraphState, or canonical ground truth.
  • User forecasts stay separate from KELLY forecasts.
  • Private watched questions stay in the browser unless a public-safe research intent is created.
  • KELLY does not treat confident text as reality. Reality closes loops.

License

KELLY and Causalea are proprietary. All rights reserved. Third-party packages and data providers keep their own licenses.

View Open Source & Third-Party Notices