A Python trading framework with a built-in multi-timeframe TA engine, historical replay and account-specific MetaTrader FX/CFD execution through MetaApi.
Explore the saved 350-day historical replay, or inspect the synthetic functionality example below. Review the recorded trades, follow the architecture and choose the modules your team wants to inspect.
Public review · No login or installation · Source/IP offer on the main site
SYNTHETIC ENGINEERING DEMOProject-generated candles, not historical market prices. The result demonstrates functionality, not trading profitability.
01 / Explore actual recorded output
A complete replay, with inspectable results.
The buyer archive was run offline on a synthetic fixture. The demo runner checked the fixture and result fingerprints, then verified the completed Journal through a read-only connection.
The dates below belong to the frozen January fixture. This page is a static view of that run, not a current market feed or a broker account.
2,304Synthetic 5-minute candles
9Plans and persistent setups
8Simulated fills and closes
PASSDeterministic demo contract
Journal COMPLETED192 derived 1h candles · 0 reported integrity violations · 100% linkage across the 8 filled signal groups
Read-only trade inspector
EURUSD · T01 · LONG
Select a recorded trade. The chart and lifecycle details update together.
Synthetic 5m closing prices · planned targets · recorded fill and exits
Exit markers use simulated account exit prices. Target levels show the stored plan; replay slippage and volume rounding can affect the recorded execution.
SIMULATED · CLOSED1h context / 5m execution
Entry fill
1.435075
Stop loss
1.428950
TP1 / TP2 / TP3
1.448000 / 1.454350 / 1.460700
Entry time
6 Jan 2026, 08:00 UTC
Final close
6 Jan 2026, 09:20 UTC
This fixture is designed to exercise the execution chain. Its price path and successful target exits are not representative investment results.
Recorded lifecycle All times UTC
Recorded partial exits
Leg
Time · UTC
Exit price
Closed share
State
Shares are of the original simulated account position. They can differ from the canonical plan fractions because of volume-step rounding. Weighted exit means the average across the closed legs, not the last target price.
Why are there 9 plans but only 8 fills?
The final opportunity remains WAITING_FOR_ENTRY with reason PERSISTENT_SETUP_CREATED at the fixture cutoff. A plan is not automatically a trade: an entry trigger is still required. The recorded snapshot contains 8 closed account executions and one unfilled setup.
Ninth opportunity: EURUSD · LONG · 1h owner timeframe · started 8 January 2026, 21:00 UTC. Input ends 9 January at 00:00 UTC, end-exclusive.
The source separates strategy/context, account policy, execution and external transport. The historical path uses a replay clock and simulated broker. The live path uses closed broker candles and the MetaApi adapter.
Select a layer to see its responsibility, source locations and a useful review question.
01 / Configuration & entrypoints
Resolve one operating configuration.
run.py starts the selected strategy runtime. falbi.py exposes supervision and read-only replay statistics. Configuration resolves operating mode, account policy and strategy settings.
Source locations
run.py falbi.py core/runtime/config.py
Review question
Are the selected modes, configuration precedence and failure messages suitable for your deployment?
Implemented starting scope
Python 3.12; MetaTrader FX/CFD through MetaApi; built-in TA Trader; HISTORY and LIVE paths; STANDARD and generic PROP policies; account-specific routing and risk; hedging/netting handling; persisted journals and recovery logic.
Integration work to plan
An independent broker API needs an adapter. Exact broker symbols/specifications must be configured. Exchange futures/options/rollover, automatic cross-broker symbol discovery, shared-write horizontal deployment and a hosted multi-tenant service need separately scoped engineering.
03 / Move from overview to source review
A concrete next step for your technical team.
The complete buyer source and operating guides are prepared. Public review materials are available immediately; controlled source inspection and complete archive delivery follow the agreed review and transaction terms.
1 / State your use case
Make the fit check specific.
Send your Python/MetaTrader stack, target instruments and account models. Indicate whether you are evaluating source/IP acquisition or another arrangement.
2 / Choose the inspection scope
Inspect the relevant implementation.
Use the source map to identify replay, TA, risk, transport or recovery modules. Ask the owner to agree a controlled source inspection and the evidence needed for your decision.
3 / Reproduce and accept
Match output to written criteria.
With the agreed buyer archive, verify the manifest, run the offline demo and inspect the Journal. Define any additional broker qualification and the delivery/rights criteria before funding.
Reproduce the offline demo after receiving the agreed source archive
Use Python 3.12 from the FALBI_1.1.0/ root. Select a fresh dependency target; do not install on top of a previous MetaApi environment. The BASE profile needs no broker account or credentials.
Expected: exit 0; DEMO | COMPLETED | status=PASS; the structural result fingerprint above; Journal COMPLETED. Inspect runtime/demo/falbi-demo-1-0-0/demo_summary.json. Read FIRST_START.TXT and DEMO.md for the exact packaged procedure.
These instructions describe the downloadable Python source edition. The public review page itself is a recorded presentation and does not run these commands.
Will a read-only broker account show the historical replay?
A broker investor login shows that account's broker-side orders, positions and history. FALBI replay trades are simulated in ReplayBroker and stored in its Journal, so they do not appear in a broker account. A separate read-only demo account would demonstrate actual broker activity, not convert replay results into broker history.
Can one strategy decision serve several accounts?
Yes, the implemented router applies a shared decision to configured eligible aliases, with independent sizing, risk and execution checks. One account may reject a trade that another accepts. The example above demonstrates one offline account, not a simultaneous multi-broker qualification.
Does this page include the full proprietary source?
The public page contains the review presentation and selected synthetic outputs. The complete source archive is provided through the agreed inspection/delivery process. Receipt or inspection alone does not transfer commercial rights, IP or redistribution rights.
What should a performance evaluation include?
Specify the exact source/version, input period and data provenance, symbols, spreads, commissions, slippage, account policy and risk assumptions. Review equity/drawdown and trade distributions alongside win rate. The synthetic example on this page is a functionality check and cannot supply market-performance evidence.