diff --git a/PUBLIC_SOURCE_MANIFEST.json b/PUBLIC_SOURCE_MANIFEST.json index 0135f4e..a5ce517 100644 --- a/PUBLIC_SOURCE_MANIFEST.json +++ b/PUBLIC_SOURCE_MANIFEST.json @@ -1,7 +1,7 @@ { "artifact_type": "dopey_public_source_manifest", "release_version": "public-source-v1", - "generated_at_utc": "2026-07-26T17:39:51Z", + "generated_at_utc": "2026-07-26T19:03:18Z", "source_repository": "ObtuseAI/dopey-private-archive", "source_commit": "a03052e787bafeb6419af440d9326fe69cf1742f", "public_repository": "ObtuseAI/dopey", @@ -17,7 +17,7 @@ "operator_agent_configuration_included": false, "aggregate_non_account_validation_metadata_included": true, "source_inventory": { - "public_file_count": 2678, + "public_file_count": 2679, "public_source_size_class": "under_50_mib", "excluded_file_count": 4514 }, diff --git a/README.md b/README.md index 0856d55..a0c7d9d 100644 --- a/README.md +++ b/README.md @@ -2,213 +2,133 @@
-DOPEY control center mark +DOPEY mark -## Evidence-first autonomous market intelligence +### Evidence-first autonomous market intelligence with a governed live boundary -**Observe the market. Challenge every claim. Learn only from settled evidence.** - -`SENSE -> MODEL -> TARGET -> ATTACK -> SETTLE -> REMEMBER -> IMPROVE -> GOVERN` +**Sense → hypothesize → falsify → settle → remember → target → govern** [![Tests](https://github.com/ObtuseAI/dopey/actions/workflows/tests.yml/badge.svg?branch=public-release)](https://github.com/ObtuseAI/dopey/actions/workflows/tests.yml) ![Python](https://img.shields.io/badge/Python-3.11%2B-3776AB?logo=python&logoColor=white) ![Desktop](https://img.shields.io/badge/Desktop-React_19-149ECA?logo=react&logoColor=white) -![Mode](https://img.shields.io/badge/operating_mode-LIVE-c2410c) -![Research](https://img.shields.io/badge/research-not_action_ready-f59e0b) +![Research](https://img.shields.io/badge/research-paper__first-f59e0b) ![Supervisor](https://img.shields.io/badge/supervisor-39_stages-8b5cf6) -![Options](https://img.shields.io/badge/options-owner_disabled-64748b) +![Options](https://img.shields.io/badge/options-owner__disabled-64748b) ![License](https://img.shields.io/badge/license-MIT-22c55e) -DOPEY desktop command center showing the governed Live Branch, research automation, evidence activity, and execution seal +
-Source-backed 1540 x 960 read-only visual-QA capture. Broker identifiers are redacted and the capture bridge has no mutation authority. +DOPEY — Evidence before promotion -### A market operating system designed to become more capable without becoming easier to fool +A market operating system designed to become more capable without becoming easier to fool. -[Operating model](#the-operating-model) · -[Loop diagram](#operating-and-learning-loop-diagram) · -[Capability diagram](#capability-topology-diagram) · -[Architecture](#system-architecture) · -[Evidence](#epistemic-honesty) · -[Quick start](#quick-start) · -[Validation](#validation-state) +[Operating model](#operating-and-learning-loop) · +[Capabilities](#capability-map) · +[Squeeze intelligence](#short-squeeze-intelligence) · +[Bubble intelligence](#bubble-leverage-ipo-and-extreme-valuation-intelligence) · +[Desktop](#native-desktop-control-center) · +[Quick start](#quick-start)
--- -## What DOPEY is - -DOPEY is a system of systems for equities, options, and macro research. It -acquires governed evidence, builds per-symbol hypotheses, attacks those -hypotheses adversarially, settles predictions against later observations, -stores the resulting memory, and uses measured outcomes to improve future -research. - -It is not a single model, scanner, strategy, or prompt loop. It is a durable -operating architecture with: - -- a point-in-time market-data spine; -- multi-horizon technical, structural, fundamental, macro, and options - intelligence; -- per-symbol targeting and specialized squeeze/bubble research; -- forward paper settlement and clustered statistical validation; -- downgrade-only epistemic supervision; -- immutable evidence, lineage, decision, incident, and mutation ledgers; -- a dependency-aware 39-stage research supervisor; -- a separately controlled Live Branch with exact paper twins. +DOPEY is an autonomous market-intelligence operating system for equities, +options, and macro research. It acquires governed evidence, builds per-symbol +hypotheses, attacks those hypotheses adversarially, records paper decisions +before outcomes are known, settles them without rewriting history, and permits +improvement only when measured fitness supports it. -The central design rule is simple: +It is ambitious about research and deliberately conservative about claims. +Backtests, replays, model prices, paper fills, and synthetic fixtures can test +mechanics; they cannot become realized or forward evidence. Missing data, +stale observations, negative-control anomalies, incomplete provenance, and +corrupt lineage fail closed. -> No claim becomes stronger merely because the system repeated it, renamed it, -> scored it confidently, or wanted it to be true. +> **Current public truth:** a fresh clone starts with no provider market data, +> credentials, broker history, account state, or private runtime lake. New +> intelligence is paper-first and reports `not_action_ready` until forward +> evidence earns a stronger state. -## The operating model +## What DOPEY does -DOPEY has two deliberately separate planes. - -| Plane | Purpose | Authority | +| Capability | What it produces | What keeps it honest | | --- | --- | --- | -| Research plane | Observe, model, rank, falsify, simulate, settle, and learn | Paper-only for new intelligence; no order construction | -| Live Branch | Reconcile the dedicated account, stage exact paper twins, review eligible actions, and record broker results | Sole broker-capable package under `dopey_live/` | - -The constitutional operating mode is `LIVE`. That does **not** mean every -research output is executable. - -- `dopey_live/` is the only broker-capable package. -- Eligible long-equity actions use the official Robinhood Trading MCP. -- Every live decision must stage an exact paper twin before broker review. -- Long-premium and defined-risk option execution is implemented but remains - owner-disabled. -- New targeting, squeeze, bubble, mutation, and epistemic surfaces are - `paper_only`, `live_blocked`, and not action ready. -- Missing authentication, provenance, freshness, identity, buying power, - reconciliation, or ledger integrity fails closed without an order. - -The authority contract lives in -[`DOPEY_CONSTITUTION.md`](DOPEY_CONSTITUTION.md). - ---- - -## Continuous workflow loops +| **Market structure** | Multi-timeframe trend, range, support/resistance, patterns, divergences, breadth, and regime context | Chronological features, point-in-time inputs, explicit contradictions | +| **Symbol research** | A joined dossier of price action, fundamentals, earnings, macro context, catalysts, liquidity, and uncertainty | Typed data gaps, freshness, and immutable source identity | +| **Short-squeeze research** | Pressure, borrow stress, settlement stress, ignition, persistence, and robustness | Open short interest stays distinct from daily short flow; pressure stays distinct from catalyst | +| **Bubble and IPO research** | Valuation and leverage fragility, IPO reflexivity, liquidity stress, and unwind confirmation | Extreme P/E or a recent IPO is never an automatic short | +| **Options research** | IV/skew/term scenarios, Greeks, contract ranking, and defined-risk paper expressions | The underlying thesis comes first; options receive no directional authority | +| **Strategy research** | Baselines, challengers, matched controls, cost stress, purged eras, and paper forecasts | Multiple-testing penalties, append-only trials, independent gates | +| **Forward learning** | Immutable predictions, later outcomes, calibration diagnostics, and abstention analysis | Event-time settlement; no replay-to-proof promotion | +| **Autonomous operations** | Dependency-aware research lanes, recovery, scheduling, and durable state | Every stage publishes its exact success, degraded, blocked, or timeout state | -DOPEY is organized as interacting evidence loops rather than a linear -"predict and trade" script. +## Operating and learning loop -### Operating and learning loop diagram +DOPEY is organized as interacting evidence loops, not a linear +“predict and trade” script. ```mermaid flowchart LR - O["1. Observe
market, macro, options, filings"] --> H["2. Hypothesize
signals, regimes, targets"] - H --> A["3. Attack
adversaries, controls, contradictions"] - A --> P["4. Paper test
costs, risk, path, exits"] - P --> S["5. Settle
forward outcomes, calibration"] - S --> M["6. Remember
immutable lineage and trade memory"] - M --> E["7. Evolve
bounded mutations and tournaments"] - E --> G["8. Govern
fitness, readiness, rollback"] + O["1. Sense
market · macro · options · filings"] --> H["2. Hypothesize
signals · regimes · targets"] + H --> A["3. Falsify
controls · contradictions · leakage"] + A --> P["4. Paper test
costs · risk · path · exits"] + P --> S["5. Settle
forward outcomes · calibration"] + S --> M["6. Remember
immutable lineage · trade memory"] + M --> E["7. Evolve
bounded mutations · tournaments"] + E --> G["8. Govern
fitness · readiness · rollback"] G --> O G -.->|"eligible evidence only"| T["Exact paper twin"] T -.->|"separate authority"| L["Live Branch"] ``` -### Loop map - | Loop | Reads | Produces | Cannot do | | --- | --- | --- | --- | -| Market sensing | Price history, options chains, macro, IPO, short-interest, margin, filings | Freshness-stamped and source-hashed observations | Invent missing data | -| Signal formation | Technical structure, regimes, volatility, flow, valuation, attention | Directional and neutral hypotheses with explicit provenance | Treat confidence as truth | -| Per-symbol targeting | Settled playbook evidence, scanner fitness, calibrated IV, liquidity | Immutable candidate registries and paper-shadow selections | Read live positions or readiness gates | -| Squeeze/bubble intelligence | Short crowding, borrow, FTD, float, valuation, IPO, leverage, options pressure | Pressure, fragility, ignition, unwind, persistence, and robustness assessments | Call crowding a squeeze or valuation a short by itself | -| Epistemic attack | Claims, sources, regime cells, denominators, contribution weights | Preserve, downgrade, block, or declare inconclusive | Promote candidates or build orders | -| Forward settlement | Paper hypotheses and later market observations | Outcome labels, calibration, clustered trust, loss attribution | Rewrite old predictions after outcomes arrive | -| Memory | Signals, entries, exits, paths, costs, regimes, failures | Content-addressed generations, graph relations, SQLite projections | Count duplicate views as independent trades | -| Evolution | Baselines, candidates, holdouts, settled fitness | Mutation lineage, tournament results, rollback requirements | Auto-promote on proxy or deteriorating fitness | -| Readiness | Independent evidence floors and reconciliation state | Exact `action_ready` or blocked status | Convert infrastructure activity into readiness | -| Live execution | Eligible action, broker state, buying power, paper twin | Queued/submitted/filled/rejected broker evidence | Infer fills, bypass the twin, or perform same-session round trips | - -### Daily research loop - -```text -Freshness and integrity - -> market and macro context - -> signals, structure, volatility, and valuation - -> per-symbol targeting - -> squeeze/bubble specialization - -> epistemic challenge - -> forward paper walk - -> settlement, trust, and calibration - -> trade memory - -> bounded scheduling and next-cycle priorities -``` +| Market sensing | Price history, options chains, macro, IPO, short-interest, margin, filings | Freshness-stamped, source-hashed observations | Invent missing data | +| Signal formation | Technical structure, regimes, volatility, flow, valuation, attention | Directional and neutral hypotheses with provenance | Treat confidence as truth | +| Per-symbol targeting | Settled playbooks, scanner fitness, calibrated IV, liquidity | Immutable candidate registries and paper-shadow selections | Read broker state or build orders | +| Squeeze/bubble intelligence | Short crowding, borrow, FTD, float, valuation, IPO, leverage, options pressure | Pressure, fragility, ignition, unwind, persistence, robustness | Call crowding a squeeze or valuation a short by itself | +| Epistemic attack | Claims, sources, regime cells, denominators, contribution weights | Preserve, downgrade, block, or declare inconclusive | Promote candidates or weaken gates | +| Forward settlement | Paper hypotheses and later observations | Outcomes, calibration, clustered trust, loss attribution | Rewrite old predictions | +| Memory and evolution | Settled evidence, failures, baselines, holdouts | Durable experience, bounded mutations, rollback requirements | Auto-promote on proxy fitness | +| Live boundary | Eligible action, reconciliation, buying power, exact paper twin | Queued, submitted, filled, or rejected broker evidence | Infer fills or bypass the twin | -Each stage publishes typed artifacts. A blocked stage remains visible as -blocked; a timeout remains `inconclusive_timeout`; a downstream consumer -cannot silently replace missing evidence with a fallback claim. +Every stage publishes typed artifacts. A blocked stage remains blocked, and a +timeout remains `inconclusive_timeout`; downstream consumers cannot silently +replace missing evidence with a fallback claim. -### Learning and mutation loop - -```mermaid -flowchart TD - B["Pinned baseline
artifact hash + measured fitness"] --> C["Candidate mutation"] - C --> V["Independent validation
holdouts + regime checks"] - V --> D{"Fitness delta"} - D -->|"decrease"| R["rollback_required"] - D -->|"equal"| K["retain baseline"] - D -->|"increase"| O["operator review only"] - R --> L["Append immutable result"] - K --> L - O --> L - L --> B -``` - -Mutation history is append-only and hash-chained. The separate head anchor -detects valid-prefix truncation. Corruption blocks future appends; the system -does not silently repair or reforge its own history. - -### Resilience loop - -The supervisor uses absorb-record-continue behavior for independent research -lanes. It records stage dependencies, start/end state, heartbeats, timeouts, -process-tree termination, recovery tombstones, disk pressure, repair debt, -and exact run identity. Independent work continues when safe; dependent work -is skipped with its causal blocker preserved. - ---- - -## Intelligence capabilities - -### Capability topology diagram +## Capability map ```mermaid flowchart TB subgraph Sources["Governed sensing"] P["Price history
corporate actions"] - O["Options chains
IV, Greeks, skew, term"] - M["Macro, rates
credit, margin, breadth"] - F["Filings, fundamentals
IPO, short interest"] + O["Options chains
IV · Greeks · skew · term"] + M["Macro · rates
credit · margin · breadth"] + F["Filings · fundamentals
IPO · short interest"] end - subgraph Core["Core market intelligence"] - R["Regime and context"] + subgraph Intelligence["Market intelligence"] + R["Regime + context"] T["Technical structure"] - V["Volatility and options"] + V["Volatility + options"] U["Fundamental valuation"] C["Catalyst identity"] end - subgraph Specialized["Specialized intelligence"] + subgraph Specialists["Specialized research"] G["Per-symbol targeting"] - S["Short-squeeze engine
pressure + ignition + persistence"] - B["Bubble engine
fragility + leverage + unwind"] - Q["Portfolio and risk
costs + concentration + stress"] + S["Short-squeeze engine
pressure · ignition · persistence"] + B["Bubble engine
fragility · leverage · unwind"] + Q["Portfolio + risk
costs · concentration · stress"] end - subgraph Truth["Truth, memory, and evolution"] + subgraph Truth["Truth · memory · evolution"] E["Epistemic supervision"] W["Forward paper settlement"] - A["Statistical and adversarial truth"] + A["Statistical + adversarial truth"] D["Durable trade memory"] - N["Bounded mutation and tournaments"] + N["Bounded mutation + tournaments"] end subgraph Authority["Governed outputs"] @@ -223,7 +143,6 @@ flowchart TB M --> R F --> U F --> C - R --> G T --> G V --> G @@ -239,7 +158,6 @@ flowchart TB G --> Q S --> Q B --> Q - G --> E S --> E B --> E @@ -254,27 +172,12 @@ flowchart TB X -.->|"independent reconciliation"| L ``` -### Capability matrix +## Intelligence built for difficult markets -| Intelligence family | What DOPEY measures | Output | -| --- | --- | --- | -| Market regime | Breadth, trend, volatility, correlation, liquidity, credit, rates, inflation, dollar, crowding | Point-in-time regime context | -| Technical structure | Multi-horizon returns, trends, breakouts, breakdowns, support/resistance, RSI, momentum, volume | Directional and neutral hypotheses | -| Statistical truth | Drift-aware baselines, event deduplication, eras, holdouts, clustered samples, deflation | Evidence-strength and fitness verdicts | -| Options | Chains, Greeks, IV, skew, term structure, gamma, spread quality, lifecycle, model repricing | Research expressions and calibrated risk | -| Fundamental/valuation | Earnings multiples, sales/book/revenue multiples, cash burn, leverage, peer/lifecycle context | Fragility and valuation context | -| IPO/reflexivity | Listing age, offer appreciation, deal size, float constraints, lock-up and overhang context | Staged IPO fragility evidence | -| Short pressure | Open short interest, float percentage, days to cover, growth, borrow scarcity, FTD stress, threshold evidence | Crowding and persistence assessment | -| Liquidity stress | Dollar/float turnover, signed volume, short flow, options turnover, gamma, borrow availability | Explicit proxy score, never hidden-intent proof | -| Catalysts/context | Filings, earnings, revisions, corporate actions, macro and news availability | Source-timed catalyst identity | -| Portfolio/risk | Exposure, concentration, liquidity, Greeks, path loss, stress ladders, kill switches | Loss-limited paper allocation | -| Trade memory | Entry/exit, path, regime, costs, P&L identity, family and cohort lineage | Reusable conditional experience | -| Adversarial intelligence | Contradictions, duplication, leakage, outliers, regime inversion, contamination | Downgrade/block/inconclusive decisions | - -### Per-symbol intelligence - -Every validated symbol remains represented even when evidence is insufficient. -The targeting lane combines: +### Per-symbol targeting + +Every validated symbol remains represented even when evidence is +insufficient. The targeting lane combines: - point-in-time history and liquidity; - current-policy playbook settlements; @@ -293,7 +196,7 @@ or readiness authority. ### Short-squeeze intelligence -DOPEY separates **crowding pressure** from **independent ignition**. +DOPEY separates **overshorted pressure** from **independent ignition**. | Layer | Evidence | | --- | --- | @@ -313,21 +216,20 @@ or weakly sourced evidence cannot qualify. ### Bubble, leverage, IPO, and extreme-valuation intelligence -DOPEY also separates **fragility** from **unwind confirmation**. - -Fragility can include: +Bubble research distinguishes an expensive or reflexive asset from a +confirmed unwind. Fragility can include: - trailing and forward P/E; -- price-to-sales and price-to-book; -- enterprise-value-to-revenue; +- price-to-sales, price-to-book, and enterprise-value-to-revenue; - valuation breadth relative to relevant history and peers; -- negative free cash flow and cash burn; -- debt, net leverage, and market-wide margin-debit pressure; -- IPO age, offer-price appreciation, deal size, float constraints, and - potential overhang; -- price extension, acceleration, attention, options crowding, IV, and gamma. - -An unwind candidate still requires independent downside evidence: negative +- negative free cash flow, cash burn, debt, and net leverage; +- market-wide margin-debit pressure; +- IPO age, offer-price appreciation, deal size, float constraints, lock-up, + and overhang context; +- extension, acceleration, attention, options crowding, IV, gamma, and + liquidity reflexivity. + +A bearish hypothesis still requires separate downside confirmation: negative returns, moving-average breaks, breakdown structure, confirming volume, put pressure, negative gamma, or a measured liquidity air pocket. @@ -341,9 +243,9 @@ a naked short or undefined-loss position. ### Observable liquidity, not imaginary omniscience -DOPEY can observe and combine borrow scarcity, float turnover, short flow, -options turnover, quote-derived gamma, and signed volume. It labels these as -liquidity-stress proxies. +DOPEY combines borrow scarcity, float turnover, daily short flow, options +turnover, quote-derived gamma, and signed volume as explicit liquidity-stress +proxies. `direct_hidden_liquidity_observed=false` @@ -351,186 +253,101 @@ Without licensed order-book, securities-lending, or participant-level data, the system does not claim to see hidden intent, dealer inventory, or a consolidated order book. ---- - -## Epistemic honesty +## Durable data and trade memory -Every targeting and squeeze/bubble claim passes through an independent, -downgrade-only epistemic stage. - -| Adversarial test | Failure caught | -| --- | --- | -| Point-in-time validation | Stale evidence, future leakage, prior-run reuse | -| Canonical fingerprints | Duplicate sources, relabeling, contribution reuse | -| Contradiction analysis | Conflicting directions within an evidence family | -| Denominator requirements | Rates and effects without enough observations | -| Contribution limits | One source or outlier dominating the conclusion | -| Regime decomposition | Simpson effects and direction reversal across regimes | -| Catalyst independence | Crowding presented as its own catalyst | -| IPO mechanics | Limited-float reflexivity presented as durable skill | -| Peer identity | Sector- or lifecycle-mismatched valuation comparisons | -| Layer isolation | Research claims contaminated by broker/order/live inputs | -| Fitness provenance | Holdout reuse, parent mismatch, proxy promotion, fitness laundering | -| Status preservation | Blocked, missing, invalid, timeout, and inconclusive states being upgraded | - -The stage can preserve a claim, downgrade it, block it, or mark it -inconclusive. It cannot promote, allocate capital, construct an order, or -change a readiness gate. - -### Evidence classes never collapse - -```text -historical observation -!= backtest result -!= replay result -!= model-priced option -!= paper signal -!= paper fill -!= submitted order -!= broker fill -!= realized P&L -``` +DOPEY treats data durability as part of model quality: -DOPEY keeps these classes separate across schemas, reports, dashboards, and -ledgers. Confidence is metadata; settlement is evidence. - ---- - -## The 39-stage supervisor - -The research supervisor is a typed dependency graph grouped into seven -functional bands. - -| Band | Stages | -| --- | --- | -| Integrity and context | Data freshness, envelope integrity, macro context, history validation | -| Signals and truth | Real signals, market structure, IV calibration, backtest, baseline arbitration, era walk-forward, statistical truth, truth-gated fitness | -| Options and fusion | Chain analytics, volatility surface, market-context fusion, straddle VRP probe | -| Forward evidence | Forward paper walk, forecast card, book quality, book risk, signal calibration | -| Adversarial learning | Adversarial stress, distributional stress, kaizen ledger, exit playbook, playbook research | -| Specialized intelligence | Symbol targeting, squeeze/bubble intelligence, epistemic supervision, adversarial panel, loss coach, target screener, universe governor, exit-policy search, chart vision, indicator foundry | -| Independent governance roots | Reliability SLO, Agentic-account reconciliation, action-readiness gate | - -`symbol_targeting` and `squeeze_bubble_intelligence` are hard producers for -`epistemic_supervision`. Account reconciliation and action readiness remain -independent roots so research cannot manufacture execution authority through -dependency pressure. +- atomic fsync-and-replace publication for critical JSON artifacts; +- manifest containment that rejects absolute paths and parent traversal; +- exchange-calendar-aware gap validation; +- finite, positive, nonnegative-volume, and OHLC-consistency checks against + active-lake bytes; +- immutable targeting-registry and trade-memory generations; +- hash-validated active pointers and rebuildable SQLite projections; +- append-only histories for attempts, findings, signals, settlements, + incidents, mutations, and live outcomes. ---- +Trade-memory outcomes use fractional model-return units and validate premium +and P&L accounting identity. Row fanout and unique economic trades are +reported separately, preventing repeated family views from masquerading as +independent experience. -## System architecture +## Resilient autonomy -```mermaid -flowchart TB - subgraph Sources["Governed sources"] - PX["OHLCV and corporate actions"] - OP["Options chains and surfaces"] - MC["Macro, rates, credit, margin"] - FU["Fundamentals, IPO, short interest"] - end +The 39-stage supervisor is dependency-aware and +**absorb → record → continue**: one failed independent research stage does +not erase the rest of the run. - subgraph Research["Research intelligence"] - SG["Signals and market structure"] - TG["Per-symbol targeting"] - SB["Squeeze and bubble engine"] - AV["Adversarial and statistical validation"] - end +It records stage dependencies, early exits, degraded states, start/end +identity, heartbeats, timeouts, process-tree termination, recovery +tombstones, disk pressure, repair debt, and exact run identity. Dependent work +is skipped with its causal blocker preserved. - subgraph Truth["Evidence and learning"] - FW["Forward paper settlement"] - TM["Trade memory and lineage"] - EV["Mutation and tournament evaluation"] - RD["Readiness gates"] - end +A transactional `SystemSnapshot` binds source and projection hashes before +cross-system decisions. Scheduling can prioritize stale or never-successful +research, but it cannot lower evidence floors, rewrite safety law, grant +broker authority, or silently enable options. - subgraph Execution["Separately governed execution"] - PT["Exact paper twin"] - LB["dopey_live / official broker MCP"] - end +Autonomy here means **durable evidence production and truthful abstention**, +not unrestricted self-modification. - PX --> SG - OP --> SG - MC --> SG - FU --> SG - SG --> TG - SG --> SB - TG --> AV - SB --> AV - AV --> FW - FW --> TM - TM --> EV - EV --> SG - FW --> RD - RD -.->|"eligible only"| PT - PT -.->|"broker review"| LB -``` +## Evidence truth -### Durable state model +DOPEY is designed to reject attractive false positives: -- Critical JSON publications use atomic fsync-and-replace. -- Manifests reject absolute paths, parent traversal, and containment escapes. -- Market-history validation checks exchange sessions, finite values, positive - prices, nonnegative volume, and OHLC consistency against active bytes. -- Candidate, memory, ledger, and mutation generations are content-addressed. -- SQLite is a rebuildable projection; immutable source generations remain - authoritative. -- Attempts, failures, rejected mutations, incidents, and live-order states are - append-only. -- Queued, submitted, partially filled, filled, rejected, and reconciled are - distinct states. +1. **No simulation laundering.** Historical replay and synthetic data never count as forward proof. +2. **No silent mutation.** Predictions and failures remain append-only; repaired candidates become new versions. +3. **No self-grading.** Candidate authors cannot promote their own work. +4. **No midpoint fiction.** Observed midpoint, model value, and conservative paper-executable price stay distinct. +5. **No complexity bonus.** A sophisticated model must beat strong simple baselines after costs. +6. **No missing-data optimism.** Stale, crossed, future-dated, or incomplete evidence forces abstention. +7. **No duplicate evidence.** Co-fired detectors and overlapping windows cannot masquerade as independent trials. +8. **No authority creep.** Automated research can tighten risk; it cannot loosen gates or unlock execution. ---- +Backtests, replay, model prices, shadow signals, paper fills, submitted orders, +broker fills, and realized P&L remain distinct evidence classes. -## Current evidence state +## Governed research and live authority -This is a source-release engineering and evidence snapshot, not a performance -claim. The underlying provider datasets and private runtime history are not -included in the public repository. +DOPEY has two deliberately separate planes. -| Surface | Latest recorded state | Interpretation | -| --- | ---: | --- | -| Validated history spine | 225 symbols / 1,808,325 bars | Broad local coverage; not proof of edge | -| Per-symbol targeting | 225 candidates / 0 verified | Current-policy settlements have not matured | -| Current-policy playbook | 572 detections / 0 duplicate appends on rerun | Fresh hypotheses, unresolved outcomes | -| Prior squeeze/bubble scan | 162 squeeze watches / 53 bubble watches | Zero action candidates in that snapshot | -| Scanner fitness v5 | 8,015 unique symbol-date events | Deduplicated mean excess over drift: -0.001436 | -| Playbook trust v4 | 0 rated families | No family meets the independent-cluster floor | -| Trade memory v3 | 321,768 rows / 186,640 unique economic trades | Durable conditional memory, not forward performance | -| Research readiness | 13 / 100 validated closes; 1 / 200 resolved decisions | `not_action_ready` | +| Plane | Purpose | Authority | +| --- | --- | --- | +| Research plane | Observe, model, rank, falsify, simulate, settle, and learn | Paper-first for new intelligence; no order construction | +| Live Branch | Reconcile the dedicated account, stage exact paper twins, review eligible actions, and record broker results | Sole broker-capable package under `dopey_live/` | -The v2 squeeze/bubble and epistemic engines in this release are implementation, -adversarial-test, and CI-gate proven. They have not been represented as a new -production outcome run or end-to-end 39-stage soak. +- `dopey_live/` is the only broker-capable package. +- Every eligible live decision must stage an exact paper twin before broker + review. +- Long-premium and defined-risk option execution is implemented but remains + owner-disabled. +- Targeting, squeeze, bubble, mutation, and epistemic surfaces remain + `paper_only`, `live_blocked`, and not action ready. +- Missing authentication, provenance, freshness, identity, buying power, + reconciliation, or ledger integrity fails closed without an order. -### Current blocking evidence +Read the authority contract in +[`DOPEY_CONSTITUTION.md`](DOPEY_CONSTITUTION.md). -1. Current-policy detections must settle before targeting can earn verified - candidates. -2. Research readiness remains below its 100-close and 200-decision floors. -3. No playbook family has 30 independent effective contested clusters. -4. Squeeze and bubble forward evidence remains insufficient for action - candidates. -5. Licensed securities-lending, richer FTD/threshold coverage, and direct - microstructure feeds remain evidence gaps. -6. Options execution remains owner-disabled. +## Native desktop control center ---- +DOPEY desktop command center showing the Live Branch, automation schedules, governed research routing, and neural evidence activity -## Desktop command center +Source-backed 1540 × 960 capture from DOPEY's deterministic, read-only +visual-QA surface. Broker identifiers are redacted and the capture bridge has +no mutation authority. -The React 19 desktop application exposes evidence and control state without -turning the UI into a second source of truth. +The React 19 desktop makes the research organism inspectable through +evidence-bound workspaces: | Market Tissue | Symbol Lab | | --- | --- | | DOPEY Market Tissue universe and regime view | DOPEY Symbol Lab options value and meta-plan view | | Universe coverage, four-tail discovery, source integrity, and regime dimensions | Per-symbol structure, scenario repricing, contract ranking, and defined-risk research | -The screenshots are deterministic visual fixtures. Model values are labeled -as model values, execution remains sealed in the capture bridge, and the UI -cannot mutate broker or research authority. - ---- +Every quantitative surface has a table, provenance view, and method +explanation. Unavailable telemetry remains unavailable; the interface does +not animate or chart fictional data. ## Repository map @@ -555,8 +372,6 @@ verifier/ protected manifests and release verification tests/ integration, safety, evidence, and release validation ``` ---- - ## Quick start DOPEY is developed and operated on Windows with PowerShell. @@ -564,10 +379,15 @@ DOPEY is developed and operated on Windows with PowerShell. ### Public data boundary A fresh public clone starts with **no provider market data, broker exports, -credentials, account history, or private runtime lake**. It includes only -source code, schemas, and deterministic synthetic/test fixtures. Operators -bring their own licensed data through governed local staging; those paths are -ignored by Git and fail closed when missing. +credentials, account history, private runtime lake, or private session +reports**. It includes source code, schemas, deterministic synthetic/test +fixtures, and one hash-pinned, test-only archive of derived paper-research +outputs. That archive contains no raw provider dataset, broker history, +current state, or forward evidence and is extracted only into isolated test +directories. + +Operators bring their own licensed data through governed local staging. Those +paths are ignored by Git and fail closed when missing. Read [`DATA_NOTICE.md`](DATA_NOTICE.md) and the machine-readable [`PUBLIC_SOURCE_MANIFEST.json`](PUBLIC_SOURCE_MANIFEST.json) before staging @@ -582,7 +402,7 @@ py -3.11 -m venv .venv .venv\Scripts\python.exe -m pip install -e ".[dev,desktop]" ``` -### Validate the checkout +Validate the checkout: ```powershell .venv\Scripts\python.exe -m ruff check . @@ -590,18 +410,7 @@ py -3.11 -m venv .venv .venv\Scripts\python.exe -m pip check ``` -The complete test suite is intentionally large. CI uses deterministic -content-addressed shards: - -```powershell -.venv\Scripts\python.exe -m operator_tools.ci_shards plan ` - --repository-root . ` - --manifest .github\ci\fast_subset.txt ` - --shards 4 ` - --output .tmp\ci-shard-plan.json -``` - -### Run bounded research surfaces +Run bounded research surfaces: ```powershell # Missing or stale governed inputs produce typed blocked states. @@ -610,7 +419,7 @@ content-addressed shards: .venv\Scripts\python.exe -m dopey_autonomy.run_resilient_research_supervisor ``` -### Build the desktop +Build the desktop command center: ```powershell cd desktop @@ -623,9 +432,7 @@ npm.cmd run build Repository commands never configure broker credentials, scrape tokens, or silently arm live execution. ---- - -## Validation state +## Public validation state | Validation | Result | | --- | ---: | @@ -638,24 +445,31 @@ silently arm live execution. | CI shard 3 | 305 passed | | CI shard 4 | 336 passed | | Repository collection | 4,334 collected / 0 collection errors | -| Final focused supervisor and epistemic integration | 31 passed | -| Ruff on changed Python surfaces | Passed | -| Python compilation | Passed | +| Focused supervisor and epistemic integration | 31 passed | | Dependency integrity | Passed | -| Workflow YAML and Bash syntax | Passed | -| Protected/live-path diff | Empty | +| Protected/live-path release diff | Empty | A repository-wide 4,332-test run reached its 15-minute boundary without a captured failure and remains `inconclusive_timeout`, not pass. The current 4,334-test collection includes two subsequent regression tests; both passed in focused and affected-shard reruns. -CI assigns each test file with `SHA-256(test path) mod 4`. Manifest reordering -or incremental additions cannot move existing files between shards. Every -assigned file must collect at least one node, node IDs must be globally -unique, and the worktree must be identical before and after each shard. +The v2 squeeze/bubble and epistemic engines are implementation, +adversarial-test, and CI-gate proven. They have not been represented as a new +production outcome run or end-to-end 39-stage soak. ---- +## Current honest blockers + +1. Current-policy detections must settle before targeting can earn verified + candidates. +2. Research readiness remains below its 100-close and 200-decision evidence + floors. +3. No playbook family has 30 independent effective contested clusters. +4. Squeeze and bubble forward evidence remains insufficient for action + candidates. +5. Licensed securities-lending, richer FTD/threshold coverage, and direct + microstructure feeds remain evidence gaps. +6. Options execution remains owner-disabled. ## What DOPEY refuses to do @@ -664,8 +478,8 @@ unique, and the worktree must be identical before and after each shard. > evidence. Substitute daily short flow for open short interest. Treat a high > P/E or recent IPO as an automatic short. Claim hidden intent from proxies. > Let a model weaken a gate. Store broker credentials. Call a queued order -> filled. Enable options silently. Claim profitability before forward -> evidence earns it. +> filled. Enable options silently. Claim profitability before forward evidence +> earns it. ## Documentation @@ -678,13 +492,13 @@ unique, and the worktree must be identical before and after each shard. - [Contributing](CONTRIBUTING.md) - [Public source manifest](PUBLIC_SOURCE_MANIFEST.json) -## License and risk +## Data, license, and risk -Code is MIT-licensed. Market-data rights remain with their providers. DOPEY -is experimental financial software, not investment advice or a promise of -returns. Options can lose the full premium paid. Live authority does not -imply model quality; operators remain responsible for independently -validating evidence, software, and risk controls. +Code is MIT-licensed. Market-data rights remain with their providers. DOPEY is +experimental financial software, not investment advice or a promise of +returns. Options can lose the full premium paid. Live authority does not imply +model quality; operators remain responsible for independently validating the +software, evidence, and risk controls.
diff --git a/docs/images/dopey_evidence_before_promotion.png b/docs/images/dopey_evidence_before_promotion.png new file mode 100644 index 0000000..f654920 Binary files /dev/null and b/docs/images/dopey_evidence_before_promotion.png differ