Research workspace · orders disabled
Research evidence
Waiting for current research evidence.
StatusAccount, services and actions that need attention
PAPER operations · No session
Your system at a glance
Loading broker, market-data, model, and risk evidence.
System snapshot—
Automatic updatesCONNECTING
SESSION STATE
Waiting for current-session evidence
Quant3 is checking the data, research, and entry schedule.
NEXT
—
PAPER AUTO NOT READY
Waiting for system evidence
Quant3 is calculating the current release gates.
0/0
checks now · not uptime
Session authorizationCheckingWaiting for current permission
New entriesCheckingEvery entry requires fresh checks
Recent stability · 15 minObservingWaiting for worker observations
Watchlist worker: checking
Account & strategy details Mode, allocation and model identity
Action queue
What needs attention
The active strategy needs a saved PAPER risk policy.
Runtime
Worker health
Account & risk
IBKR PAPER snapshot
Starlight IB Gateway
Gateway and Quant3 share a loopback-only PAPER API. Broker credentials and MFA are entered only through the SSH-tunneled operator desktop.
Requires the Windows launcher on this computer.
Finish the Windows launcher update
- Find install-quant3-windows-launcher.ps1 in your browser’s Downloads. Downloading does not install it.
- Right-click the Quant3 script and choose Run with PowerShell. Review any browser or Windows warning privately; this page cannot approve or bypass it.
- Return here and choose Open desktop. Allow the external application prompt if shown. You can also open MaveSense Quant3 Desktop from the Windows Start menu.
The updater preserves an existing verified WSL launcher. Desktop passwords stay private.
Desktop did not open?
Allow the browser’s external application prompt if shown. If the browser never hands over to Windows, open MaveSense Quant3 Desktop from the Windows Start menu. Check that its secure connection completed. The browser cannot confirm whether the viewer opened.
Reopen an existing tunnel’s viewerKeep your SSH, VNC and IBKR passwords and MFA private.
Release gates
Safety checklist
Release gates
Safety checklist
Manual control
Mode & emergency actions
Operations
Freshness & recovery
Exceptions
Open incidents
0 selected
| Select | Source | Severity | Status | Category | Summary | Opened | Action |
|---|
No visible incidents.
Page 1 of 1
Study collection alerts
Research workspace
Research
Daily scans, rejection reasons, market context and the data collection queue.
Today’s workflow
Data → decision → execution
Market-data ingestion0%
No ingestion heartbeat is available.
1Market dataWaiting
→
2Research scanWaiting
→
3Candidates0 selected
→
4PAPER orders0 queued
Next research actionWaiting for readiness evidence.
Market regime
UNKNOWN
0/100Composite
Waiting for complete daily market evidence.
Awaiting current analysis.
Risk off0–39Neutral40–69Risk on70–100
Strategy risk budget0.00×No new exposure
Intraday shock guardUNKNOWNAwaiting status
US 10-year Treasury
—Context onlyAwaiting published yield data.
Daily publication · does not change the regime score.
Source: U.S. Treasury
Evidence cutoffUnavailable
CalculatedUnavailable
Completed daily bars
Decision funnel
Latest session decision paths
Click a filter box to configure its settings.
Screening preview with current settings
Preview counts include every earlier filter. The recorded strict scan and active intraday watchlist have separate counts and source times.
Independent setup checks
Each check covers the retained stocks separately. These counts do not add up to selected trades.
Why stocks were rejected
A stock can have several rejection reasons.
PAPER_VALIDATION overlay
Eligible is the full daily pool. Up to 100 candidates are watched, with strict selections taking priority; SPY is collected separately. Evaluated counts distinct candidates, not repeated five-minute checks.
Daily screen + intraday monitor
Scanner evidence
MONITOR UNKNOWNNO DAILY RUN
S&P members scanned—
Retained for setup tests—
V1 evaluation coverage— / —
Monitor-pool eligible—— strict selected · — rebound eligible · — active capped set
No completed scan evidence is available.
Every retained symbol receives a V1 evaluation score. Setup gates, the strict score gate, and the PAPER_VALIDATION overlay remain separate.
| Symbol | Outcome | Progress | V1 evaluation | Strict score gate | Overlay priority | Liquidity rank | Evidence as of | Why |
|---|
No symbols reached this scan scope.
0 decisions
Page 1 of 1
Research inputs
Data collection
Extra historical collection LOADING
History and company calendars collect outside market-data hours. Source availability is tracked separately from verified company listing dates.
Collection issues & automatic retries
Next window is the earliest permitted collection time, not a guaranteed dispatch. Shared credits and other jobs can delay it. A provider start date is not a verified IPO date.
| Company | Dataset | Requested range | Reason / source boundary | Attempts | Last attempt | Next window |
|---|
Market context
Reference prices only. Every PAPER entry requires a new eligible execution quote no more than 2 seconds old.
Rates · daily publication
US 10-year Treasury yield
Awaiting published yield data.
Yield date and last checked will appear here. This is a published daily yield, not a live quote.
Market intelligence
Top market-impact headlines
No qualifying source-backed market headline was published in the last 36 hours.
Recent & upcoming releases
Economic indicators
No recent or upcoming economic release is available.
Company research
Companies
Daily technical charts, financial statements, earnings and linked news for S&P 500 companies and your saved watchlist stocks.
Find potential bottoming stocks · draft research
Formation research · draft model
Potential bottoming stocks
Spot a decline near a recent low, then look for price stabilization, RSI recovery and easing MACD selling pressure. Confirmation requires a later close above the potential signal’s high. A break below its reference low invalidates the setup.
Run a scan to assess stored daily data.
| Stock | Formation / data | Score | Potential detected | Reference low | Price session |
|---|
—
Scores rank formations; they are not win probabilities. This version is unqualified and does not change current entries or the frozen study. A new prospective validation is required before promotion. Earnings, news, liquidity and execution checks remain separate.
Select a company
Daily technical chart
NOT LOADED
Last daily close Not loadedInspected date Not selected
Detector validation record
Research observations are saved to the server for your account when automatic validation is On.
EMA 20EMA 50Candles & volumeRSI 14 · MACD 12/26/9
Signal key · P / R bottoms · T / D tops · × invalidated
Potential bottom / confirmed rebound appear below candles. Potential top / confirmed downturn appear above candles. Markers show detection sessions, not hindsight extremes. Hover for a preview; click a candle, marker or use the date slider to pin inspection.
Inspected signals · draft research
Choose a stock to inspect signals.
Full detector evidence & validation
Chart source & freshness
Choose any company to open its stored daily bars.
Company matches from the configured general feeds; an empty list does not prove that no company news exists.
Data dates and history availability
Company calendar
Earnings in the next five trading days
Awaiting the earnings calendar
Upcoming earnings are flagged by company. Weekends and US exchange holidays are excluded; today's releases are shown separately.
| Symbol | Company | Release date | Flag | Release time |
|---|
Missing-issuer clearance queue
| Symbol | Company | Required session | Reason | Retained source | Receipt expiry | Responsible owner | Next action |
|---|
Company research sources
Company data coverage
Awaiting fundamentals status
| Dataset | Fresh companies | Attempted | Needs attention | Latest attempt |
|---|
Awaiting company-news coverage
News sources and collection receipts
Collection schedule unverified. The current configured source status is shown above; archived financials are not a promised refresh.
Experimental PAPER
One strategy per session
Declare separate trials for strict baseline, rebound overlay and staged pullback. A plan grants no authority. Each session requires separate PAPER_VALIDATION authorization and an original pre-open capture before its arm can be authorized. These experiments do not promote a model.
No plan loaded. Complete stocks may proceed only when their evidence and shared account checks pass.
Declare trial dates & budgets
Model workspace
Model
Frozen rules, evaluation plans, performance evidence and promotion review.
Your strategy uses quantitative rules, not a trained machine-learning model. The selection score ranks setups; it is not a win probability. Research detectors have their own Research section and require separate validation before becoming trading rules.
Prospective research
Automatic formation evidence
Waiting for automatic capture evidence.
Trial admission evidence not loaded.
Universe, missing inputs and outcome receipts
| Session | Declared | Usable | Unavailable | Pending capture | Deadline | Measured | Source gaps | Price-basis gaps | State |
|---|
Price-unit evidence review queue
Queue not loaded.
| Symbol | Original session | Outcome horizon | Review state | Owner | Next action |
|---|
The export is an incomplete request, not an admission. Retain authentic source bytes and verify original and later price units, then check a completed receipt with the existing owner-reviewed tool. Missing evidence stays unmeasured.
Passive recorded execution
Execution quality evidence
No measured sample loaded.
Timing, costs and limitations
| Application timing | Samples | Median ms | 95th percentile ms |
|---|
Draft research · automatic validation
Improve bottom and top detection
Qualification schedule not loaded. Refresh research evidence; owner-reviewed price units are required.
Compare the original detector with two fixed candidates: a stable base with momentum agreement, and a stricter price structure break. Declare one untouched prospective cohort before outcomes exist. Select using its training period, then require later validation and forward confirmation. Missing sources or insufficient evidence keep the original.
Only future formations after a locked trial passes use its validated profile. Past signals are not moved to hindsight extremes. These per-company diagnostics do not promote a trading model or qualify the historical backtest.
Draft model · quant3-potential-bottom@0.1.0
Bottoming formation detector
On · research detector enabled. This setting does not enable trading.
Oversold waiting → potential bottom → confirmed rebound. Price, RSI and MACD are evaluated on completed daily bars. Later prices cannot move an earlier signal to the lowest candle.
Starts after a 5% decline near a 20-session low with RSI at that low at or below 40. A potential setup needs price recovery plus improving momentum. Confirmation requires RSI recovery and a later close above the potential candle’s high. The reference low and expiry limit define failure.
Experimental thresholds need prospective validation and cost testing. Historical chart markers are reconstructed from the current adjusted series, not point-in-time performance evidence. This detector is separate from the active strategy and frozen study.
Draft model · quant3-potential-top@0.1.0
Topping formation detector
On · research detector enabled. This setting does not enable trading.
Overbought waiting → potential top → confirmed downturn. Starts after a 5% rise near a 20-session high with RSI at that high at or above 60. A potential setup needs price weakness and RSI rollover, weakening MACD or bearish RSI divergence.
Confirmation requires a later close below the potential candle’s low, RSI rollover and two decreases in the MACD histogram. A later high above the reference high invalidates the setup. Waiting setups expire after 15 sessions; confirmed setups after 10.
Experimental thresholds require separate validation. Completed daily candles only. Historical markers are reconstructed research, not admitted performance evidence. The detector does not authorize selling or shorting.
Research study
Strategy comparison
Compare three fixed strategies using the same costs and risk limits. Simulated research only; this study sends no orders.
Not loaded
Awaiting study
No progress loadedLoad the frozen study or preview its design.
Next step
Refresh the study to check its progress.
Results so far
Results are preliminary. Scores rank setups, not win probabilities. Holdout results stay sealed until the final session ends.
Study rules & schedule
- Strict baseline: existing V1 selections, up to 15 stocks and two per sector.
- Rebound overlay: existing non-selected rebound candidates, up to 20 by their current priority.
- Staged pullback: admit RSI 30–45 formations after a prior close above the 50-day EMA. Retain existing watchlist members for at most five subsequent completed sessions; fill vacancies with new formations by RSI ascending, then symbol, capped at 20. Confirm later with at least three RSI points of recovery and a close above the previous daily high. Current event and liquidity checks apply.
- All arms then require the existing five-minute entry trigger, daily-ATR stop sizing, 1.5R target, session forced exit and portfolio limits. Different candidate caps are part of the registered strategies.
The design fixes 20 future training sessions, 20 validation sessions and 40 holdout sessions. Holdout outcomes remain sealed until the final session ends. A semantic correction requires a separately audited successor; the original record and sealed holdout remain retained. It sends no orders.
Measurements & source-gap evidence
Full results & export
| Partition | Strategy | Sessions | Trades | Net P&L | Expectancy (R) | Profit factor | P&L at double costs |
|---|
Missing pre-open captures are not replayed as prospective evidence. Five-minute marked drawdown, stop rejections and post-exit observations are simulation diagnostics; they cannot prove broker execution quality. Historical archives still require dated source admission in the evaluation section below.
No study loaded.
News evidence
Unresolved news & issuer coverage
Review exact headline revisions and retrieval gaps. Proposals are retained for a separately versioned policy; they do not unblock current news decisions or change the frozen study.
Current systemic news gate has not been checked.
Active entry-news coverage has not been checked.
Corrected policy preview is inactive.
Load news evidence.
Issuer retrieval coverage
Each row describes a bounded discovery attempt. A verified publication date is not complete coverage; an empty search is not proof of no material news.
| Issuer | Last observed | Retrieval | Verified / discovered | Next attempt | Coverage |
|---|
Historical evaluation
Evaluation progress
Check dated inputs and recorded sessions for the current implementation.
No plan loaded.
Source coverage and model diagnostics
| Required authentic source | Missing sessions | Dates | Owner | Next action |
|---|
Missing historical outcomes remain unmeasured. Continue the separate prospective study while obtaining dated archives; elapsed future sessions cannot be manufactured.
Inspect session requirements
Missing membership and news counts overlap. File integrity checks do not admit an archive into replay. These rows show the exact dates and input failures still needing work.
| Session | Partition | State / requirement | Last input failure | Retry eligibility |
|---|
Check the plan to see runnable sessions. A request can wait for permitted hours for up to 24 hours. Once execution begins, its 15-minute budget limits starting additional sessions; an in-progress session finishes safely.
A completed scan with no setups is valid. Missing inputs are not a zero-trade result. Model results require the separate performance verifier.
Active strategy only
Promotion evidence
Counts apply only to the active strategy version. Synthetic and replay artifacts are shown as non-promotional and never satisfy a promotion gate.
Evaluation evidence
Model testing progress
- Collection
- Awaiting the data worker
Company data and history are collected on weekends, exchange holidays and outside the pre-open and regular session. Completed price windows still need dated membership and corporate-action validation for credible backtesting. A completed scan can validly select zero stocks.
Versioned PAPER controls
Portfolio policy & pass conditions
Define capital, risk, and reserve intentions while seeing the tighter rules the active strategy actually enforces.
Policy version
—
Not loaded
Requested allocation is not an execution override.Loading active strategy constraints…
Independent model review
Promotion path & evidence
Next controlled transition—
Required evidence—
Reviewer authority—
Advanced evidence and promotion review
Inspect prerequisites before opening reviewer actions. These controls require independent reviewer authority.
Requested investedEmergency reserve— headroom
Loading order limits…
Risk configuration
Portfolio variables
Edit portfolio variables
Rule reference
Frozen V1 baseline & risk-policy limits
FROZEN V1RISK POLICYMONITOR ONLYPROHIBITED
Setup, ranking, universe, entry and exit rows describe the frozen V1 baseline. Risk rows show the requested policy and V1 hard limits. Saved current-session gate settings are a separate layer; historical results retain their recorded rules.
Current-session gate settings have not been verified. Inspect Research for the effective saved settings; these baseline rows do not validate edited gates.
Current Quant3 PAPER portfolioPAPER · SIMULATED CAPITAL
Portfolio performance
Read-only portfolio report · no account snapshot loaded.
Select Portfolio to load the current PAPER account. This report does not change trading authority.
Broker account NAV
—Whole PAPER account · awaiting sourceBroker account cash
—Whole PAPER account · awaiting sourceTracked gross exposure
—Quant3 strategy positions onlyOpen position P&L
—Unrealized · tracked positions onlyRecorded closed-trade P&L
Quant3 strategy results by exit date, in US dollars. Open-position P&L and broker account valuation are separate.
Recorded net results already include recorded costs. Deposits and withdrawals are not classified as investment gains. These figures are not time-weighted investment returns.
Account valuation history
Observed broker net liquidation value. Includes external cash flows; not an investment-return chart.
View valuation observations
| Observation | NAV | Data quality |
|---|
Cash & concentration
Account cash and tracked holdings are different scopes.
Tracked holdings concentration
Weights use gross market value of complete, priced strategy positions only. They do not reconcile the full broker account or imply available buying power.
Stock holdings
Current positions recorded by Quant3 for this PAPER account. Other broker holdings may be outside this scope.
| Mark quality |
|---|
Monthly profit record
Recorded closed-trade net P&L. Current month is month-to-date.
| Period | Net P&L | Closed trades | Coverage |
|---|
Calendar-year record
Current year is year-to-date. Missing history is not a zero return.
| Year | Net P&L | Closed trades | Coverage |
|---|
Recent closed trades
Recorded strategy activity. These rows are historical records, not trade instructions.
| Stock | Closed | Recorded net P&L | Cost evidence | Exit |
|---|
Methodology & data quality
Quant3 PAPER simulates trading. This report is not a live investment track record or a permission to trade. Operating mode and entry authority remain controlled by the existing system safeguards.
Time-weighted and money-weighted returns are unavailable without a complete external cash-flow ledger. No percentage investment return is inferred from changes in account value.
CSV contains the full returned report, independent of the holdings search. Missing values remain blank and retain their coverage labels. Printed reports preserve the as-of time and scope.
Execution records
Trading
Orders, fills, positions and the reasons behind each recorded decision.
Your stocks
Watchlists
Loading watchlists…
Search any supported US-listed stock. On follows your current strategy and model; Off keeps it saved without watchlist-driven collection or evaluation. Every PAPER entry still requires ready data and all trading checks.
| Watchlist | Stock | Strategy | Evaluation status | Latest five-minute bar | Actions |
|---|
Point-in-time evidence
Trading lifecycle
Stored decisions, risk, broker state, protection, fills, and exits. No values are reconstructed in the browser.
No lifecycle evidence matches this view.
0 records
Session archive
Reports
Published session results and evidence quality. Current collection progress lives in Research.
Immutable daily evidence
Session reports
Zero-trade sessions remain first-class evidence. Checksums and provenance are shown with every report.
Archive
Daily reports
Reports have not been loaded.
No published reports yet
A report appears after the scheduled session report job publishes its retained evidence. This empty archive does not prove a completed or failed trading session.
Next: review the collection workflow for its verified schedule or missing inputs, then refresh this archive after publication.
Report detail
Select a session
No report has loaded.
Choose a report to inspect its assessment, funnel, operational evidence, and provenance.
Single-session evidence
Why no order was created
Trace one real market session from its immutable daily candidate ancestry through persisted intraday, risk, and broker evidence. Overlapping reasons are reported separately from mutually exclusive states.
Choose a session date to load its read-only rejection evidence.
Immutable evidence path
Session decision funnel
Overlapping reason evidence
| Scope | Reason | Occurrences | Distinct candidates | Rate |
|---|
No reason evidence was recorded for this session.
Mutually exclusive state evidence
| Scope | State | Occurrences | Distinct candidates | Rate |
|---|
No state evidence was recorded for this session.
Detected bottlenecks
No bottleneck code was detected.
Interpretation limits
No additional limitations were reported.
SUPER_ADMIN · LATEST-SCAN WHAT-IF ONLY
Parameter testing lab.
Compare bounded daily-setup and monitoring thresholds against the latest immutable research scan.
Test-only profiles never place orders or change the active strategy.
Preview re-evaluates stored scan evidence. Saving creates an immutable what-if version; it does not activate, promote, tune, or authorize any parameter.
Before running a what-if
Checking preview prerequisites
Bounded experiment
What-if parameters
Latest immutable scan
Baseline vs what-if
| Symbol | RSI | Setup baseline | Setup what-if | Monitored baseline | Monitored what-if | What changed / test reasons |
|---|
Run Preview to compare the current fields with the frozen latest-scan baseline.
Fixed multi-session sensitivity
20-session RSI scenario replay
NOT LOADEDNON-PROMOTIONAL
Compare the frozen V1 baseline with four bounded RSI scenarios: maximum 50, two rising comparisons, two recovery points, and the combined 50 / 2 / 2 case. Results measure daily-setup sensitivity, not projected trades.
Choose a through-session date to run the read-only sensitivity replay.
Aggregate scenario comparison
| Scenario | Changed thresholds | Baseline threshold passes | Scenario threshold passes | Added | Removed | Pass rate | Sessions with additions |
|---|
No scenario results are available.
Combined 50 / 2 / 2 by session
| Session | Evaluated | Baseline threshold passes | Combined threshold passes | Added | Removed | Added symbols | Removed symbols |
|---|
No combined-scenario session results are available.
Fixed safeguards
No safeguards were returned.
Evidence limits / requirements
Minimum evidence is satisfied and no additional limitation was reported.
Append-only record
Saved what-if versions
| Version | Label | Created | Created by | Reason | Hash |
|---|
No test-only version has been saved for this strategy.
System administration
Administration
Access, security, audit history and database maintenance.
Database evidence integrity
SUPER_ADMIN
Identity & security
Deactivation, role changes, MFA resets, and session revocation are audited. Historical records cannot be deleted.
Enrollment
Create user
Shown once. Transfer through a secure channel.
Authorization
Users and sessions
| User | Roles | Status | Sessions | Actions |
|---|
Authentication
Login history
| Time | User | Outcome | Source | Client |
|---|
Loading…
Tamper-evident trail
Privileged and audit actions
| Sequence | Time | Role | Action | Target | Reason | Event hash |
|---|
Loading…