Project page · technology by RYIUK
// DAAVILE FX — flagship project · finance pillar inside RYIUK

DAAVILE FX. Backtested capital, gate-first.

The finance side of RYIUK's proposition — first hostile domain for the stack. What matters here is the sauce: evidence ledger, multi-era truth, staged gauntlet, execution realism, ML oracle, agent fleet, knowledge flywheel. Methodology is the moat; the trading is just where it runs.

§ 01 — methodology · the sauce
Seven pillars

The algorithm is the hypothesis.
The pipeline is the answer.

10,000+ verified backtests, an untouched out-of-sample era, Monte-Carlo stress gates, and a machine-learning oracle that reads the internals to decide which candidate deserves silicon next. This is the moat — methodology, not marketing. Operator-grade secrets stay private; the discipline is public.

Fail any gate → blocked. Nothing is declared; things survive. Receipts over rhetoric · staged promotion · append-only truth.
01 · Evidence ledger

Every claim has a run identity

Append-only ledger of real MT5 backtest runs. Deterministic, pattern-seeded run names; epoch rebases when the market-data base changes. Every champion number traces to a re-runnable run — nothing is anecdote.

append-onlyrun identityepoch rebase
02 · Multi-era truth tables

Regimes, not a single scoreboard

Hand-curated date ranges spanning genuinely independent market regimes — institutional era, consolidation, modern tick era. Plus a completely untouched out-of-sample holdout no optimization has ever seen: the final judge.

independent erasuntouched holdout
03 · Robustness gauntlet

Staged promotion, fail = blocked

Champions are survivors. Same ladder for every candidate: clean → latency (150/300/500ms + jitter) → in-EA Monte-Carlo lab (tick noise, skipped fills, spread bursts, lot jitter, exit noise) → cost stress → forward paper.

latency ladderMonte-Carlocost stressforward
04 · Execution realism

Latency as a distribution

Per-range gates: strict profitability under delay, survival under worse delay, a ceiling, and a jitter hard-veto. Ranking by probability-weighted expectation per range — never averaged into one fake global number.

E[metric]per-rangejitter veto
05 · ML oracle

The machine that reads the machine

Per-range surrogates predict outcome before a burn; feature importance surfaces which internal logics actually drive results; dead-knob detection kills wasted sweeps; Bayesian + low-discrepancy mass sweeps generate candidates; active learning ranks the next teachable experiment.

surrogatesdead-knob detectionactive learning
06 · Agent fleet operations

A backtest factory

Parallel MT5 terminals through a typed MCP protocol, orchestrator manifests, campaign chunking, owed-cell recovery so mid-flight kills resume cleanly, terminal-ops discipline — one driver per manifest, orchestrated like a data center.

typed MCPmanifestsclean resume
07 · Institutional memory · knowledge flywheel

The compounding asset

Structured plain-text knowledge base — internal logic map, per-range learnings, distilled lessons, champion doctrine — RAG-queryable by the agents today, explicitly designed as training fuel for a future LLM fine-tune. Today it's the agents' shared memory; tomorrow it's the model.

RAG for agentsper-range learningsfine-tune fuel
// flagship project · campaign numbers
16 / 16
pass bar per gauntlet · no partial credit
5,000
prospects screened daily · against the full bar
+200
ranges in measurement · thousands of windows · 20+ years
840
config dimensions explored · 20 × 21 × 2
~10×
screen vs confirm tier · measured
3,681 / 7,941
logic-level rows / real deals in models
§ 02 — DAAVILE FX · flagship project
The champion gate

16/16, or it dies.
One gauntlet is a unit. We run thousands.

Flagship project of DAAVILE FX — the finance pillar inside RYIUK. The multi-window backtest gauntlet that decides which configs survive. The rest of the lab feeds it and ships beside it.

One gauntlet = sixteen fixed windows across regimes — crash, bears, drift, the long slog, live 2026 slices. Same setfile. Fixed lot. Risk-management levers banned from the search. Pass means every window — no partial credit. The campaign runs daily R16 epochs like it's nothing: thousands of windows, measurement blocks past +200 ranges, 5,000 prospects a day against the full bar.

PASS requires positive profit in the window, and profit greater than max equity drawdown. Fail once → the candidate is documented, not promoted.
2011 crash 2012–2015 bears 2016–2019 drift 2004–2015 slog 6 × 2026 slices
Standing champion GODTST3_j245_g6x2475
Pass bar 16 / 16 · no partial credit
Lineage bet_1580→MUSETST-R3_w01g2005→GODTST3_j245_g6x2475
Config 840 dims · fixed lot · no RM levers
Span 20+ years of price · bulls · bears · crashes · chops
Throughput 5,000 prospects/day · daily R16 epochs · +200-range blocks

The scoreboard is public because the failures are too. Growth isn't “everything works” — it's “we kept less than we generated, on purpose.”

§ 03 — DAAVILE FX · execution flow
The gate pipeline

Hypothesis in. Ledger entry out.

How the gate project runs one candidate from idea to verdict: a documented key change per setfile. A cheap screen culls what can't beat round numbers. Real-tick confirmation is the only run that counts. Then it faces the champion — or the kill ledger.

01STATE
Config genome
840-length vector · every bit a knob the engine reads
→
02MUTATE
Documented hypothesis
one key change per setfile · rationale logged
→
03SCREEN
Tier-1 fast pass
~15.4s / window · cheap cull before real compute
→
04CONFIRM
Real-tick gauntlet
~152.1s / window · tier-2 · only this tier writes the ledger
→
05GATE
Champion or kill
beat on all 16 → promoted · else → kill ledger, reasons attached
§ 04 — DAAVILE FX · receipts
The kill ledger

Ideas are cheap.
Dead ones keep the pipeline honest.

We might not be a quant shop. We still score like one. Recent kills — each logged, dated, kept.

Flagship honesty · the epoch rebase

We found our own data had drifted

Mid-campaign, the pre-2016 tick history on the terminals was found drifted. Everything built on it was invalid. The wrong move is usually convenient: keep the numbers.

kill ledger · restoration pending

A second kill entry is missing

This section originally held two kill entries. During the dedupe refactor only one was captured into capabilities.json and the other was dropped from the markup. Nothing was invented to fill the gap: the entry is restored from a copy of the page or left deleted, deliberately.

§ 05 — machine learning
ML, with the same discipline

Models score like the strategies do

The evidence discipline we run against strategies applies to every model in the stack — feature stores from verified runs, honest validation, nothing promoted on vibes.

measured · 2026-10-12 · leave-one-era-out

A model could pick the winning config on an era it never saw

HistGB over the config matrix: top-1 pick accuracy 50% across held-out eras (random = ~4%), profit capture 0.79, picked config ≈ 2.7× the range's mean outcome. That's the receipt for "distill the algorithm."

measured · 2026-10-12 · smoke gate

The decision layer accepted the EA's records end to end

32 range-choice records → kev option ranking on Qwen3-0.6B (LoRA-16, CPU): 0 rejected, saved clean; one epoch in-sample 37.5% vs 18.75% zero-shot. Small numbers — the fine-tune gate says more labels first (~100 outcomes), so this stays a mechanics proof, not a claim.

design · receipts-bound

The algorithm stays code. The model learns its judgment.

Executor deterministic → kev-style policy ranks the options it already computes → resident 0.6B explains via RAG and abstains when no receipt matches. The LLM never sits in the tick path.

§ 06 — decision layer
kev · in evaluation

Even the tools get gated

A small packed-question decision model — Jev-style — sits in the loop as an experiment judge. Isolated per-question forward passes, each calibrated, answering things like: “violates a known dead lever?” — “targets a window regime this stack never opens?”

The standard for every tool: not “does it sound clever,” but “does it survive the harness.”
Shadow verification Livescores every proposed mutation · no authority (yet)
Claim QA Liveexperiment claims cross-checked against run records
Base featurization Killedby the frozen harness
Similarity ranking Killedby the frozen harness
Fine-tune gate Gatedomain training only at ~100 labeled outcomes — until then, shadow
§ 07 — system map
The ecosystem

One stack. Every app talks to the others.

Dashboards, lab tools and backtesting engines built to connect — the LLM handles backend, configs and GUI automation; you operate from one place. Status tags reflect reality: live, in progress, or research.

multi-symbol

One account. Every symbol.
The thing MT5 cannot do.

MT5 tests one symbol at a time, with its own margin. Our portfolio engine runs N symbols through one shared account — one wallet, one margin pool, one equity curve — so a gold position and a euro position fight for the same free margin the way they would in a real account.

Shared margin, real fight

A gold position eats margin the euro position wanted. The gate refuses the second trade and counts the refusal — never a silent no-op. Pending orders reserve the pool before they fill.

Shared gateReservationHonest refusals

Real algo DLLs, one account

The same C-ABI DLLs the terminals run drive each leg, but they see the portfolio balance and their opens hit the shared gate — the algo knows the account is shared, because it is.

C-ABIPortfolio vtableShared balance

Walk-forward + Monte Carlo

Every window is a full merged session judged by the gauntlet band law. The Monte Carlo shuffles the session's closed trades and reports p5/p50/p95 balance, drawdown and survival rate — deterministic under seed.

Band-law windowsLCG shuffleSurvival rate

conditional statistics

Ask the data a question.
Get N and a confidence interval, not a vibe.

P(outcome | conditions) over your own bars, with the sample size and the 95% Wilson interval on every number. The law is the same everywhere: no code path prints a percentage without its sample size.

gapFill WHERE dayOfWeek = Tue

estimate 52.1%  N = 685  95% CI [48.4%, 55.8%]
stability: 56.1% vs 48.1% · halves agree
per-year: 2024 58.9% · 2025 48.1% · 2026 47.9%

How often gold fills its opening gap, on your data, with the denominator honest and the split-half stability on the same line.

Wilson 95%Halves-stabilityN guards

orbBreak(15, up) WHERE session = newyork

estimate 51.0%  N = 664  95% CI [47.3%, 54.8%]
stability: 50.3% vs 51.8% · halves agree
per-year: 2024 51.2% · 2025 52.6% · 2026 48.5%

Opening-range breaks in the New York session. Every outcome composes with every condition — day-of-week, gap size, month, anything the registry can express.

ORBgap filtersPer-year

the law as a library

The gauntlet, distilled to pure Rust.
Band law. Gates. Ranges. Protocol schema.

The same laws that gate every strategy — now a standalone crate that anything in the ecosystem can call: the band law (G/Y/O/R/X/U/D with the p/dd ratio ladder), the 16-cell range canon, the promotion gates, the protocol schema validator. One engine, one discipline, every surface.

Band law

Worst-band fusion, never an average: G ≥ p/dd 1.0, Y ≥ 0.5, O < 0.5, R < 0, X = blown, U = not run, D = can't run. A blown leg vetoes — a huge drawdown on a loss reads grey, not red.

Worst-bandBlown-firstp/dd ladder

Range law

16 canonical cells + OOS windows, content-hashed and immutable. Unique ids, real calendar dates, 10-year span cap, from-before-to, every phase in a stage — the structure a run binds to.

R1–R8Content-hashedImmutable

Protocol schema

Versioned JSON with strict validation: unknown keys are rejected, not dropped — a typo'd field surfaces as an issue, never a silent no-op. Schema v2 with MT5 bindings (symbol, modelling, deposit, leverage, timeframe) and sub-template references.

Strict documentMT5 bindingsTypos rejected

robustness

Stress every axis at once.
One broken corner breaks the claim.

The full cross-product grid — adverse slippage × spread-burst multipliers — each cell a full merged session judged by the band law. The verdict fuses by the worst band across the whole grid, never by the average.

Slippage × spread grid

0/1/3/5 points of slippage crossed with ×1/×3/×10 spread widening. A strategy that only works at zero slippage and the real spread reads fragile, and the report says so.

Cross-productWorst-band fusionSpread-burst

10× spread cannot help

The spread-burst axis widens every tick symmetrically around the mid before it reaches the broker, the strategy, or any algo DLL — the whole session trades a burst-spread market. Deterministic, no RNG.

Symmetric wideningAlgo sees itDeterministic

live + local

The agents have eyes now.
Quotes. Your bars. The TA kernels.

Keyless live quotes for the whole basket, the vault's own bars on disk, and the same indicator kernels the strategy interpreter runs — all exposed to agents through MCP tools, no keys required for the core.

Keyless quotes

Gold, the Dow, the Nasdaq, the S&P, bitcoin, and four FX pairs — all mapped, all keyless, all live. The agents can ask "what is gold doing right now?" and get the price with a day range and a change %.

XAUUSDUS30BTCNo keys

Your bars, your vault

65 GB of tick history and every bar file under the vault — the last N bars in M1 or D1, read from the .qidx sidecar without touching the network. The data is yours; the read is local.

Local vault.qidx indexedM1 / D1

The same TA kernels

SMA, EMA, Wilder RSI, MACD(12,26,9), Bollinger(20,2σ), ATR(14) — computed on the same bars with the same definitions the strategy interpreter runs, so the agent's read and the engine's read agree.

Wilder RSIMACDATRBollinger

in build · gated ladder · kev's successor

Don't fine-tune the model first.
Teach the pipeline. Then earn the weights.

The distill program is the rung above kev's shadow lane: an explicit gate ladder from imitation to a state-conditioned selection policy that must beat the always-on champion. Every rung is scored by one yardstick on the same forward split, against a seeded noise floor — and no GPU is spent before a cheap local gate says the signal exists.

static best 0.682 · typical config 0.390 · ridge-on-params 0.421 · oracle 1.150
noise floor p95 0.481 · recency +0.169 · pf-gate +0.163 · state-knn +0.104
matrix 250 cells · 113 configs · 16 windows · OOS 2020–2025 sealed · tracked at daavfx/daavilefx-distill

The ladder, not the model

G0 data integrity is live and passing. G1a is a kill. G1b carries the first signal. G2 — "does the LLM add value over classical on held-out windows" — is the program's thesis gate, and the LLM is formally blocked until the matrix says it has something to learn from.

G0–G5One yardstickNoise floor

The G1a kill

Ranking configs by their parameters scores 0.421 — below doing nothing clever (0.390) and far below the static best (0.682), beating it in 7% of windows. Parameters carry no forward signal. The selection program does.

KilledLeak-free split14 windows

The epoch audit

839 of 4,033 ledger rows — 35% of what posed as ground truth — were a superseded epoch measured on retired terminals. The anchor's 2.05 score was irreproducible: source setfile gone, EA version unrecorded. Legacy is now excluded by default and every exclusion is counted.

839 rowsSupersede-never-eraseAuditable

Harness revisions

Thirteen single-knob selection programs scored against the seeded random floor: recency +0.169, pf-gate +0.163, min-obs +0.159, state-knn +0.104; ridge blends −0.23. Honest caveat baked into the report: at 14 windows these are hypotheses, not conclusions — the sweep confirms or kills them.

One knobSame splitHypotheses

Market state, leak-free

Trailing features per window from the 478 MB bars vault, one streaming pass — computed only from bars before the window starts, so a crash inside the window cannot change its features (regression-tested). The four windows born at the vault's edge carry honest nulls. R16 trails bear: vol90 0.35, dd90 0.19.

Anti-lookaheadTrailing onlyHonest nulls

The queue

Cross-sweep manifests are built — K=30 and K=50, 480/800 runs across all 16 windows, resumable, holdout mechanically refused. Results fold into a v2 ledger the moment they land. The LM enters as a kev-style revision proposer with calibrated abstention; weights only after it beats classical.

K×16Operator-gatedG2 pending
§ 08 — work queue
Right now

What we're actually building

No fake launch badges. This is the live queue across the whole lab — local LLM harness, Rust parity, finished apps, model tiering, pattern mining, and the gate project's open windows.

01
Gate project: push measurement past +200 ranges

Daily R16 epochs are routine — the queue is wider regimes, more ranges, same 16/16 bar, zero partial credit.

02
Local LLM + harness for agentic backtesting pipelines

llama.cpp driving the full MT5 backtesting flow end to end.

03
Rust parity with MT5 — bridge tests, Quantum cross-checks

Same results as the terminal, at Rust speed.

04
Hermes & OpenCode as MT5 backtesting helpers

Agents operate terminals, register champions, stop only at verified beats over the champion.

05
Quantum & Premium: agentic + GUI visual backtesting

All tools, skills and scripts RAG-backed; loops built in.

06
Finish the Lab apps & both dashboards

Copy Trader, MT5 MCP, Data Downloader, Prop Firm Lab, Risk Manager, Quantum, Premium.

07
Model tiering: what to fine-tune, what stays RAG

1B–2B executors · 2B–4B auditors · larger MoE for main audits. Trained free on Colab/Kaggle.

08
Pattern mining across bars, ticks & setfiles

Move-based features — it moved $X, not "it was $2000 in 2020" — feeding strategies, LoRA data, risk profiles.

09
Distill the algorithm into a small local decision model

Executor stays deterministic; a gated ladder decides when a model earns its place. Live now: G0 integrity PASS, G1a KILLED parameter-conditioning (0.421 vs 0.682 static — parameters carry no forward signal), G1b state-conditioning first signal (+0.104 above the noise floor). Next: the K×16 cross-sweep densifies the matrix, then a kev-style revision proposer — weights only after it beats classical on held-out windows. Free-tier only. Tracked at daavfx/daavilefx-distill.