Position lifecycle semantics and stop loss review suite

Trading reports explain position rules through the same typed LiSA Script contracts that validate them, with seven labeled strategies covering six canonical stop loss families and one managed hybrid.

LiSA Script

• Unified expression validation and lifecycle wording through one typed contract for function arguments, result domains, historical offsets, lifecycle availability, and human readable meaning.
• Added donchian(period, "lower"|"upper"[, offset]) for lookahead safe channel boundaries from complete genuine candle windows.
• Removed the separate lifecycle expression parser, duplicate function lists, and compatibility formatting paths.

Backtesting engine

• Kept one generated Position lifecycle companion for every open position action, with Strategy JSON as the raw executable reference.
• Added isolated fixed, ATR volatility, previous Donchian channel, RSI thesis invalidation, favorable extreme trailing, and completed candle time exit comparisons using the same Supertrend entries and execution assumptions.
• Kept the managed position strategy as a seventh hybrid comparison with a partial target, break even protection, ATR trailing, and an opposing Supertrend exit.
• Preserved the financial results of all five existing Trading strategies while expanding the staged catalog and report evidence.

Strategy Lab

• Centralized DCA and Trading labels through one backend registry that combines authored classification and study labels with indicators derived from validated strategy expressions.
• Added Trading label filters and a dedicated stop-loss-suite label for reviewing the seven comparison strategies together.
• Expanded the Trading catalog to nine 4h strategies: seven review suite members plus the Supertrend line and deterministic timed diagnostics.

Managed-position reporting

Trading reports present authored management rules and each position lifecycle as backend-prepared report content.

Backtesting engine

• Added position-opening frequency from executed position lifecycles and the complete effective backtest duration.
• Classified the protection active at a final exit as the initial, break-even, or managed stop without changing execution results.

Strategy Lab

• Added compact managed-position instructions to Trading decision-tree actions and clearer lifecycle labels to chart markers and position outcomes.
• Grouped each closed position with its opening, partial target fills, material stop changes, final closure, fees, and total result.
• Moved Win rate into the Strategy visualizer and added Trades per day to Execution activity in place of the final-exit-mix KPI.

Managed Trading positions

Trading positions can close fixed portions at targets and tighten their shared stop after target fills or completed candles.

LiSA Script

• Added named entry-time value snapshots, ordered targets sized as percentages of initial executed quantity, and one-time or recurring tighten-only stop updates.
• Kept the existing take-profit shorthand as one full-position target and retained position-bound indicator exits.
• Added position reads for entry price, genuine completed-candle age, initial stop distance, favorable extrema, and named entry values.

Backtesting engine

• Added partial position exits within one flat-to-flat trade, including proportional realized PnL, per-fill fees, remaining quantity, trigger cancellation, and protection history.
• Made the shared stop close the current remaining quantity after any target fill; target allocations continue to use initial executed quantity.
• Activated fill-driven stop changes from the next genuine minute and completed-candle changes from the first genuine minute after that close.
• Kept conservative stop-first handling for unresolved same-minute active triggers and atomic deterministic partial position exits without an order-book liquidity model.

Trading Lab

• Added a managed-position Supertrend comparison beside the unchanged fixed-protection strategy.
• Added backend-prepared partial target markers, segmented stop rails, target rails, fill quantities, and lifecycle evidence.

Supertrend and position-bound exits

Trading strategies can use Supertrend states and bind named completed-candle exit conditions to an opened position.

LiSA Script

• Added supertrend(period, factor) with explicit bullish and bearish state reads, historical offsets, and a matching indicator declaration.
• Added optional ordered exit_conditions groups inline on open_position actions.
• Removed the strategy-specific seeded-random and fixed-cadence expression helpers.

Backtesting engine

• Added TradingView-style Supertrend calculation from Wilder ATR and completed candles, with no future-candle reads.
• Evaluated position-bound exit conditions while their position is open and executed a selected exit at the next genuine candle open.
• Kept take-profit and stop-loss execution ahead of pending condition exits, including entry-candle and next-open gap handling.

Strategy Lab

• Replaced the seeded ATR showcase with a 4h Supertrend long-short strategy while retaining its 2% planned stop-risk sizing and ATR 14 protection.
• Added a backend-colored bullish and bearish Supertrend overlay, compact authored exit policies in entry details, and condition traces for position exits.

Trading risk sizing, fees, and report metrics

Trading Lab can size positions from planned stop risk and reports backend-calculated account, risk, trade, and execution metrics.

LiSA Script

• Added optional risk_percent_equity_at_stop sizing beside the existing equity-percentage position sizing. Only the ATR strategy opts into the new mode.
• Kept explicit 1x leverage for equity-percentage actions and recorded the effective leverage produced by every executed position.

Backtesting engine

• Added position sizing that targets a selected planned stop-price loss against positive pre-entry account equity, using the actual next-open entry and frozen stop level.
• Applied a fixed 0.045% fee to entry notional and exit notional. Fees remain outside the planned stop-risk target.
• Added marked leverage, exposure, drawdown, return, benchmark, trade-quality, duration, and fee metrics with explicit unavailable and unbounded states.
• Kept leverage uncapped for this testing phase without collateral, maintenance-margin, liquidation, or funding simulation.

Strategy Lab

• Replaced Trading KPI placeholders with backend-owned display values across the summary, account growth, visualizer, and execution sections.
• Changed Trading Execution activity to one position-opening category, with UTC-day counts and backend-formatted direction and action details.
• Improved Trading chart inspection with decision traces for strategy actions and lifecycle details for positions.

Historical indicator reads and ATR protection

LiSA Script can read current or historical indicator values and use resolved indicator expressions for Trading position protection.

LiSA Script

• Added one explicit trailing historical offset contract across registered technical indicators, price, time, and numeric external-series reads.
• Required indicator calculation parameters in executable expressions and declarations instead of relying on implicit formula defaults.
• Added atr(period) with any positive integer period, complete-window initialization, and Wilder smoothing.

Trading Lab

• Added entry-derived take-profit and stop-loss distances from general numeric expressions, including current or historical indicator values.
• Resolved protection distances on the completed signal candle, applied absolute levels from the actual next-open entry, and kept those levels fixed through position close.
• Added a deterministic 4h ATR long-short stress strategy using one fixed 17-genuine-candle schedule and stable seeded direction choices.

Trading position execution charts

Trading Lab now shows complete position lifecycles and requires take-profit and stop-loss protection on every opened position.

Trading Lab

• Added backend-prepared position spans with long and short duration tint, entry rails, exact-price execution points, action markers, and position details.
• Required entry-relative take-profit and stop-loss levels on every Trading open action while retaining conditional full-position close actions.
• Resolved daily and 4h protection exits against canonical Kraken 1-minute candles, including entry-candle exits and genuine opens beyond a level.
• Applied a conservative stop-loss-first result when one minute crosses both configured levels.
• Kept protection rails visible from position entry through take-profit, stop-loss, or a separate strategy exit.
• Removed adverse Trading slippage adjustments so actions use the exact modeled fill price while fees remain separate.
• Consolidated the current Trading catalog on one deterministic 4h long-short lifecycle strategy.

4h Trading Lab and Kraken market history

Trading Lab supports a separate 4h BTC comparison group and a rebuilt Kraken-only historical market database.

Trading Lab

• Added the 4h Trading timeframe and isolated comparisons by timeframe and fixed period.
• Evaluated signals after completed 4h candles and executed selected actions at the next tradable 4h candle open.

Data / DB

• Rebuilt Trading history from Kraken BTC/USD spot data only, separate from the DCA database and stored Trading results.
• Stored one canonical sparse 1-minute history through July 28, 2026 and derived UTC-aligned 4h and daily candles, checked against available native Kraken interval files.
• Extended the archive history from Kraken PostTrade executions after an exhaustive 129,509-candle comparison matched the official first-quarter 2026 OHLCVT release.
• Kept official Kraken source omissions out of storage while using explicit nontradable previous-close slots for regular 4h indicator timing.
• Added checksum-pinned staging, source and coverage audits, atomic publication, and rollback backup handling for Trading market updates.
• Updated the internal Market Data page with per-series coverage, construction, source omissions, validation, and build status.

Daily Trading Lab

Hey-LiSA now separates DCA research from daily BTC long and short strategy research.

DCA Lab and Trading Lab

• Added separate top-level DCA Lab and Trading Lab areas.
• Kept existing DCA strategies, accounting, stored results, and reports unchanged.
• Added the first daily BTC long and short Trading research path.

LiSA Script

• Added the explicit trading category alongside dca.
• Added account, position_state(), open_position, and close_position.
• Limited the current Trading contract to one position, long or short, at 1x leverage.
• Required an explicit close before opening in another direction, with no increase, reduction, partial close, reversal, or automatic flip.

Trading Lab

• Added Trading-specific cards and reports for account equity, return, PnL, drawdown, fees, trades, and current position.
• Added backend-prepared Trading charts, action markers, legends, tooltips, trade details, and decision trees.
• Added daily signal evaluation at completed candle close with execution at the next candle open.

Data / DB

• Added a dedicated Trading BTC database, separate from the existing Kraken DCA database.
• Added internal Market Data visibility for coverage, provenance, gaps, freshness, and refresh status.

Minor fixes and improvements

Strategy Lab

• Ranked visible Strategy Lab and Test Bench strategies by score after label filtering, including the force-visible DCA baseline.
• Added backend-defined indicator bands and reference levels, LiSA Script-derived strategy thresholds, and specific series and component names to the indicator legend below the chart.
• Added a backend-owned strategy-action legend above charts, including display-ready action labels, sizing, marker colors, shapes, ordering, and deduplication.
• Added a strategy-period initial viewport with duration-relative margins, retained warmup and context data, and preserved zoom and pan.

LiSA Script

• Added public ath() for the highest available-history BTC closing price through the current candle, including chart overlay support and public documentation.

Data / DB

• Added transactional refresh-reports tooling for rebuilding stored Strategy Lab and Test Bench card and report payloads from existing results without rerunning simulations.

Spot sell support in Strategy Lab

LiSA Script and the local backtesting engine now support simulated BTC spot sells for portfolio strategies.

LiSA Script

• Added a simulated sell action for BTC spot strategies.
• Added sell sizing for fixed BTC amount, percent of current BTC position, and full current BTC position.
• Limited sells to spot behavior: sell actions can reduce existing simulated BTC holdings but cannot create short exposure, leverage, margin, or perp behavior.

Backtesting engine

• Added sell execution beside the existing buy path, with BTC position caps, quote cash proceeds, fees, slippage, and no negative BTC balances.
• Added sell accounting fields to stored result summaries, including BTC sold, quote proceeds, realized PnL, and remaining cost basis.
• Added a normalized RSI14 spot buy/sell DCA example strategy and stored results across the preset Bitcoin periods.

Strategy Lab

• Updated backend-owned report payloads so trade markers, tooltips, execution maps, and decision-tree action details distinguish buy and sell orders.
• Updated Strategy Visualizer display so sell markers render as sell-side events while the frontend continues to read backend-prepared payload fields.

Strategy labels and LiSA Script normalization

Strategy labels are now first-class in LiSA Script, with label filtering across Strategy Lab and Test Bench, plus a cleaner public strategy language.

Strategy labels

• Added required, validated strategy classification labels.
• Added consistent strategy-label display and filtering to Strategy Lab and Test Bench.
• The normalized DCA baseline now remains visible under Strategy Lab filters without assigning it labels it does not have.

LiSA Script

• Replaced separate baseline moving-average tools with one generic ma() primitive supporting mode, period, and timeframe parameters.
• Required moving averages to use a full lookback window, including available candles before the selected backtest period, and marked runs unavailable when the required history does not exist.
• Removed authored rule, branch, IF, action, and condition IDs from LiSA Script; the backend now generates deterministic structural IDs.

Strategy Lab

• Consolidated the Strategy Decision Tree into one canonical visualization and removed the duplicate tree.

Small Strategy Lab and docs fixes

A focused follow-up for clearer Strategy Lab reports, synchronized chart panes, normalized labels, and updated LiSA Script docs.

Strategy Lab

• Added a backend-owned Portfolio Growth milestone annotation for the last major portfolio value milestone reached during a selected period, shared by Strategy Lab and Test Bench reports.
• Updated Strategy Visualizer charts so indicator panes share the same X/time axis as the price chart.
• Normalized strategy card and report labels to focus on market, strategy style, and indicators used instead of noisy configuration fragments.

LiSA Script

• Updated the public indicator and expression docs to match the current built-in technical indicator catalog.
• Fixed the indicator reference table so expression reads and quoted component names render correctly.

Data / DB

• Aligned the Rego BTC DCA30 Test Bench window and related stored report data to the December 26, 2016 start date used by the reconstruction.

Technical indicator platform

A descriptor-driven indicator platform for LiSA Script, with broader built-in technical indicators, example strategies, and cleaner chart panes in Strategy Lab.

LiSA Script

• Moved technical indicators into one catalog package per indicator so new indicators can be added without growing one shared indicator file.
• Added built-in technical indicators for SMA, EMA, MACD, RSI, Bollinger Bands, Stochastic, ADX, Parabolic SAR, CCI, Williams %R, and Ichimoku.
• Added scalar expression reads for multi-output indicators, including MACD components, Bollinger bands, Stochastic lines, ADX / DI lines, and Ichimoku components.
• Indicator formulas remain internal, with no TA-Lib or other technical-analysis runtime dependency.

Strategy Lab

• Updated Strategy Visualizer panes so technical indicators, external series, and state variable lines stay separated by backend-owned chart payloads.
• Added support for multi-series oscillator panes, histogram series, point overlays, band overlays, and cloud-style overlays.
• Added one normalized 50 quote/day example strategy for each new indicator while keeping the existing RSI example strategy.

Backtesting engine

• Added per-backtest indicator caching so repeated indicator reads do not recompute full indicator series on every candle.
• Added targeted stored runs and result checks for the new indicator example strategies across the preset Bitcoin periods.

Data / DB

• Stored the new indicator strategy results with the same canonical card and report payloads used by the rest of Strategy Lab.

LiSA Script cleanup and Strategy Lab tooling

A cleaner LiSA Script layer, stronger Strategy Lab reports, safer scoped backtest iteration, and more coherent curated strategy data.

Strategy Lab

• Improved Strategy Detail reports with clearer Portfolio Growth metrics, return-on-deployed-capital context, fuller DCA asset-delta comparisons, and backend-owned report data.
• Improved Strategy Visualizer review with backend-owned decision traces, readable action details, action-colored markers, and cleaner bounded external-series panes.
• Revamped Decision Tree review with clearer readable rule, condition, branch, and action details in Strategy Detail.
• Added backend-provided data-size indicators to Strategy Lab cards so large detail reports are visible before opening them.

LiSA Script

• Published and expanded the public LiSA Script reference for the implemented JSON strategy language: rules, branches, grouped conditions, expressions, actions, indicators, variables, external series, constraints, and execution behavior.
• Strengthened user-facing strategy nodes with stable machine IDs and readable names, including first-class action IDs and action names.
• Expanded builtin indicator work with generic moving-average expressions, weekly intervals, multiplier overlays, and less strategy specific language.
• Added state-indicator and post-buy state variable patterns for plotting strategy-owned dynamic values.

Backtesting engine

• Improved scoped backtest iteration with targeted precompute, stored-result updates, clearer progress output, and behavior guards.
• Moved more report and visualizer preparation into backend-owned payloads so frontend report components render prepared data instead of inferring strategy semantics.

Data / DB

• Cleaned curated strategy data and external indicator coverage while pruning stale experimental fixtures.
• Strategy Lab cards stay lightweight while full selected-report payloads are stored for detail pages.

Strategy Lab reports and stored result data

Cleaner Strategy Detail reports, stable stored result data, and lighter cards for browsing backtest results.

Strategy Lab

• Rebuilt Strategy Detail as a unified report with top summary, portfolio growth, strategy visualizer, daily execution map, cash reserve, decision tree, and access to the strategy's LiSA Script source.
• Added interactive Lightweight Charts sections for portfolio growth, cash reserve, and strategy visualizer views.
• Added a Cal Heatmap daily execution map with strategy action colors and period navigation.
• Integrated the strategy graph, equity comparison, Strategy Visualizer, and LiSA Script source inspection into the report.
• Added Trade / State inspection in Strategy Lab for reviewing trade decisions, before/after state, and execution details.

LiSA Script

• Added evaluated LiSA Script decision traces to Strategy Lab reports so trade review can show which rules, branches, and conditions fired.
• Prepared report payloads around stable strategy-language IDs for decision trees, trade tooltips, LiSA Script source access, and visualizer data.

Backtesting engine

• Added backend Trade / State inspection data for before/after state and execution details.
• Protected strategy-result outputs with result snapshots, checksums, schema tests, and payload tests.
• Cleaned product API boundaries so Strategy Lab consumes prepared report data instead of raw backtest result blobs.

Data / DB

• Added stored report data for Strategy Detail pages.
• Added lightweight stored card data for Strategy Lab catalog cards and Test Bench cards.

Strategy Lab browser

Browsable Strategy Lab catalog, result cards, preset Bitcoin market windows, early Strategy Detail reports, and custom Test Bench cases inside the app.

Strategy Lab

• Expanded the curated strategy catalog around Daily DCA, cash baseline, drawdown strategies, moving-average strategies, CBBI, Fear & Greed, Pi Cycle, and stateful examples.
• Improved Strategy Lab cards with period rows, baseline comparison markers, portfolio composition, return/deployment metrics, and sparklines.
• Added early Strategy Detail reports for deeper result review.
• Added Test Bench cases inside Strategy Lab for custom periods, custom options, and DCA comparison workflows.

LiSA Script

• Added specific PoC built-in tools for moving-average strategies, including ma(...), highest_close(...), Pi Cycle, and 2-Year MA logic, before a fully normalized indicator layer.
• Added PoC external-series semantics with external_series declarations and series(id) reads to prove CBBI and Fear & Greed strategies.
• Expanded stateful strategy examples using variables, conditions, branches, and action rules; this was not an exhaustive LiSA Script standard, and later cleanup and normalization are still expected.

Backtesting engine

• Added cash-only baseline comparisons for Strategy Lab result review.
• Updated fixed-today boundaries and period names for reproducible backtest comparisons.

Data / DB

• Added preset Bitcoin market windows including cycle-top ranges, ETF cycle, and first 100K cycle.
• Added halving markers plus external-series import, storage, and coverage support for Strategy Lab comparisons.

Also added the public SvelteKit website for docs, HQ, contact, Foundry, bilingual routes, and changelog.

Strategy Lab PoC

First working proof-of-concept for Bitcoin strategy research and backtesting.

LiSA Script

• Designed LiSA Script, our custom in-house JSON-based strategy language for defining conditions, actions, branches, indicators, variables, external series, and simulated execution rules.

Backtesting engine

• Added the FastAPI app and CLI entry points for running strategy research workflows.
• Built the first Python backtesting engine with cashflows, simulated orders, equity curves, portfolio summary, and drawdown/return analytics.

Strategy Lab

• Designed the first Strategy Lab App in Svelte for browsing strategies, selecting periods, and inspecting stored backtest results.
• Added early charts, portfolio summaries, order/trade inspection, strategy graphs, and equity comparisons.
• Added the first Daily DCA 50 baseline strategy.

Data / DB

• Added SQLite storage for strategies, candles, and stored backtest results.
• Added BTC candle backfill, Kraken daily market-data ingestion, and candle repository support.
• Added repeatable checks around market-data import, stored backtest results, and LiSA Script behavior.