mt5cli

Generic MT5 data and execution infrastructure for Python applications. Export from the CLI or import a small, stable Python API in downstream packages.

The Public API Contract lists stable SDK exports (mt5cli.STABLE_SDK_EXPORTS), CLI commands, internal helpers, and responsibilities that remain out of scope (strategy logic, backtests, optimization).

Built on top of pdmt5, a pandas-based data handler for MetaTrader 5.

Architecture

• pdmt5 — canonical MT5 client, DataFrame/trading primitives, and MT5 constant parsing (TIMEFRAME_*, COPY_TICKS_*, order types).
• mt5cli — public MT5Client API, standardized dataset schemas, storage helpers, CLI commands, and SQLite history collection built on pdmt5.
• mt5api — sibling HTTP adapter for remote MT5 access; not a dependency of mt5cli.

Features

• Multi-format export: CSV, JSON, Parquet, and SQLite3 output formats
• Auto-detection: Format detection from file extensions
• Comprehensive data access: Rates, ticks, account info, symbols, orders, positions, and trading history
• Flexible timeframes: Named timeframes (M1, H1, D1, etc.) and numeric values
• Connection management: Optional credentials, server, and timeout configuration
• SQLite rate loading: Load mt5cli-managed rate tables/views for offline workflows

Installation

pip install -U mt5cli MetaTrader5

Parquet export is not included by default. To enable it, install the parquet extra:

pip install -U "mt5cli[parquet]" MetaTrader5

Python API (downstream packages)

Import MT5Client for generic MT5 data access, schema normalization, and optional order primitives.

from datetime import UTC, datetime
from pathlib import Path

from mt5cli import (
    MT5Client,
    build_config,
    collect_history,
    mt5_session,
    update_history_with_config,
)
from mt5cli.schemas import DataKind, normalize_dataframe
from mt5cli.utils import Dataset, export_dataframe

# Persistent session for multiple calls
with mt5_session(build_config(login=12345, server="Broker-Demo")) as client:
    rates = client.copy_rates_range(
        "EURUSD",
        timeframe="H1",
        date_from="2024-01-01",
        date_to="2024-02-01",
    )
    positions = client.positions()
    check = client.order_check({"action": 1, "symbol": "EURUSD", "volume": 0.1})

# Normalize MT5 frames to the public schema contract before storage
closed_rates = normalize_dataframe(
    rates, DataKind.rates, symbol="EURUSD", timeframe="H1"
)
export_dataframe(closed_rates, Path("rates.csv"), "csv")

# Bulk SQLite history (same behavior as collect-history CLI command)
collect_history(
    Path("history.db"),
    symbols=["EURUSD"],
    date_from=datetime(2024, 1, 1, tzinfo=UTC),
    date_to=datetime(2024, 2, 1, tzinfo=UTC),
    datasets={Dataset.rates, Dataset.history_deals},
)

# Incremental append for automated pipelines
update_history_with_config(
    output="history.db",
    symbols=["EURUSD"],
    config=build_config(login=12345),
)

Schema contracts live in mt5cli.schemas (DataKind, validate_schema, normalize_dataframe). Export and storage helpers are in mt5cli.utils (Dataset, export_dataframe) and mt5cli.history.

MT5Client.order_send() is a live execution primitive: it can place real trades on the connected account. mt5cli does not implement strategy logic, signal generation, backtesting, or optimization — downstream applications must gate live execution explicitly.

Trading lifecycle and state helpers

Trading applications can depend on mt5cli imports only; terminal path, credentials, server, and timeout are forwarded to pdmt5.Mt5Config, numeric login strings are coerced to integers, and empty login strings are treated as unset. Pass allow_whole_dollar_env=True to expand ${ENV_VAR} and bare $ENV_NAME placeholders in connection string parameters before coercion.

from mt5cli import (
    build_config,
    calculate_spread_ratio,
    get_account_snapshot,
    mt5_session,
)

# Login from environment — numeric string is coerced to int automatically
config = build_config(login="$MT5_LOGIN", allow_whole_dollar_env=True)

with mt5_session(config) as client:
    account = get_account_snapshot(client)
    spread = calculate_spread_ratio(client, "EURUSD")
    positions = client.positions(symbol="EURUSD")

CLI usage

# Export account information to CSV
mt5cli -o account.csv account-info

# Export EURUSD M1 rates to Parquet
mt5cli -o rates.parquet rates-from --symbol EURUSD --timeframe M1 \
  --date-from 2024-01-01 --count 1000

# Export ticks to JSON
mt5cli -o ticks.json ticks-from --symbol EURUSD \
  --date-from 2024-01-01 --count 500 --flags ALL

# Export symbols to SQLite3 with custom table name
mt5cli -o data.db --table symbols symbols --group "*USD*"

# Export with connection credentials from env or placeholders
MT5_LOGIN=12345 MT5_PASSWORD=secret MT5_SERVER=MyBroker-Demo \
  mt5cli -o positions.csv positions
MT5_PATH="/path/to/terminal64.exe" mt5cli -o positions.csv positions

Run as a Python module:

python -m mt5cli -o account.csv account-info

Commands

Use order-check to validate a request payload before running order-send --yes. close-positions is the safer high-level alternative that builds correct close requests automatically. At least one --symbol or --ticket must be provided. CLI connection flags fall back to MT5_LOGIN, MT5_PASSWORD, MT5_SERVER, and MT5_PATH when unset, and explicit CLI values still win.

history-gaps

Inspect collected SQLite rate views offline and export one row per detected gap. For managed compatibility views, the command infers bar spacing from the view name. Use --granularity-seconds for custom tables or views.

mt5cli -o gaps.json history-gaps --sqlite3 history.db
mt5cli -o eurusd.csv history-gaps --sqlite3 history.db --table rate_EURUSD__M1_1

collect-history

Collect several historical datasets per symbol into one SQLite database in a single MT5 session. Pick datasets with repeatable --dataset (default: rates, history-orders, history-deals; add --dataset ticks when tick-level history is required — tick data can grow the SQLite database quickly), choose conflict behavior with --if-exists append|replace|fail (default: fail), and optionally derive cash_events / positions_reconstructed views from history_deals via --with-views.

mt5cli -o history.db collect-history \
  --symbol EURUSD --symbol GBPUSD \
  --date-from 2024-01-01 --date-to 2024-02-01 \
  --dataset rates --dataset history-deals \
  --timeframe M1 --flags ALL --if-exists append --with-views

History orders and deals are fetched per symbol and concatenated, so the symbol filter is applied consistently across all datasets. The cash_events view is derived from symbol-filtered history_deals, so account-level cash events with empty or non-matching symbols may be excluded. The rates table records the requested timeframe so appended runs at different timeframes remain distinguishable. The positions_reconstructed view aggregates trade deals by position_id, excludes positions without closing-side entries, and uses volume-weighted open/close prices; reversal deals (DEAL_ENTRY_INOUT) are reported via volume_reversal / reversal_count columns.

Grafana-ready SQLite dashboards

mt5cli can prepare a SQLite database for use as a Grafana datasource (via the SQLite plugin or similar). Most grafana_* views expose an integer epoch-second time column for use in Grafana time-series panels. Two views (grafana_realized_pnl, grafana_trade_stats) are static symbol-level summaries with no time column — use them in table or stat panels.

Prepare the schema (idempotent, no MT5 connection needed)

mt5cli -o history.db grafana-schema

This creates snapshot tables (account_snapshots, position_snapshots, order_snapshots, terminal_snapshots, snapshot_runs) and all grafana_* views and indexes in the SQLite database. Safe to run repeatedly — all operations are idempotent.

Snapshot current account state

mt5cli -o history.db snapshot \
  --symbol JP225 --symbol HK50 --symbol NL25 \
  --with-account --with-positions --with-orders --with-terminal \
  --with-grafana-schema

Appends one timestamped row per data type. Never places orders or modifies trading state. Run periodically (e.g. from a cron job or a loop) to build a time-series account history.

SDK usage

from pdmt5 import Mt5Config
from mt5cli import update_observability_with_config

# This wrapper owns the MT5 connection automatically.
update_observability_with_config(
    output="history.db",
    config=Mt5Config(login=12345),
    symbols=["EURUSD", "GBPUSD"],  # optional position/order filter
    include_account=True,
    include_positions=True,
    include_orders=True,
    include_terminal=True,
    with_grafana_schema=True,
)

Available Grafana views

Time-series views (integer epoch-second time column; snapshot views also expose run_id):

Static summary views (no time column; use in table or stat panels, not time-series):

Example Grafana queries

-- Equity curve over time
SELECT time, equity FROM grafana_account_snapshots ORDER BY time;

-- Rolling balance by account login
SELECT time, login, balance FROM grafana_account_snapshots
WHERE login = $login ORDER BY time;

-- Open positions at latest successful snapshot
SELECT symbol, volume, profit FROM grafana_position_snapshots
WHERE run_id = (SELECT MAX(run_id) FROM snapshot_runs WHERE status = 'ok');

-- Realized PnL by symbol
SELECT symbol, total_profit FROM grafana_trade_stats ORDER BY total_profit DESC;

Grafana and telemetry API docs

The shipped Grafana helpers are documented in docs/api/grafana.md, including publish_grafana_copy() for creating a WAL-safe published SQLite copy for Grafana.

OpenTelemetry metrics are documented in docs/api/telemetry.md, including enable_otel_metrics(), configure_metrics(), the mt5cli[otel] extra, and OTEL_EXPORTER_OTLP_ENDPOINT.

Incremental history SDK

For automated pipelines, use the importable incremental API instead of re-fetching fixed date ranges:

from pdmt5 import Mt5Config
from mt5cli import update_history_with_config
from mt5cli.utils import Dataset

# This wrapper opens and closes MT5 for you.
update_history_with_config(
    output="history.db",
    symbols=["EURUSD", "GBPUSD"],
    config=Mt5Config(login=12345),
    datasets={Dataset.rates, Dataset.history_deals},
    timeframes=["M1", "H1"],  # default: all fixed MT5 timeframes
    lookback_hours=24,
    create_rate_views=True,
    with_views=True,
    include_account_events=True,
)

• collect-history: explicit date-range export into SQLite.
• update_history: incremental append based on existing SQLite MAX(time) per symbol (and timeframe for rates); account-level deals use a separate cursor when include_account_events=True.
• rates table: normalized storage with symbol and timeframe columns.
• Rate compatibility views: mt5cli manages all rate_* views. Naming is rate_<symbol>__<timeframe> when a symbol has one timeframe, otherwise rate_<symbol>__<granularity>_<timeframe> (for example rate_EURUSD__M1_1). Stale rate_* views are dropped and recreated when rates change for offline downstream tools.
• Rate view resolution: use resolve_rate_view_name() / resolve_rate_view_names() to map symbols and granularities to existing SQLite compatibility views without creating databases. Both accept None (or a missing path) and return deterministic default names unless require_existing=True.
• Rate view loading: use load_rate_data() / load_rate_data_from_connection() to load a SQLite rate table or view into a DatetimeIndex DataFrame.
• Multi-series rate loading: use build_rate_targets() to build neutral RateTarget(symbol, timeframe) pairs, resolve_rate_tables() to map them to table/view names (pass require_existing=True for strict resolution), and load_rate_series_from_sqlite() to load them into a mapping keyed by (symbol, integer timeframe). The loader requires existing managed views unless explicit_tables is supplied, and rejects duplicate (symbol, timeframe) targets.
• Multi-account latest rates: use collect_latest_rates_for_accounts() with AccountSpec to read the latest bars for several account groups, merged into a (symbol, integer timeframe) mapping. For long-running pollers, collect_latest_rates_for_accounts_with_retries() adds bounded exponential backoff that retries only recoverable MT5 errors and re-raises once retry_count is exhausted.
• Latest closed bars: use collect_latest_closed_rates_for_accounts() when downstream logic must exclude the still-forming current bar. It fetches count + 1 bars at start_pos=0, drops the last row with drop_forming_rate_bar(), and validates each series is non-empty. collect_latest_closed_rates_by_granularity() returns the same data keyed by (symbol, granularity_name) such as ("EURUSD", "M1").

from mt5cli import AccountSpec, collect_latest_closed_rates_by_granularity

rates = collect_latest_closed_rates_by_granularity(
    [AccountSpec(symbols=["EURUSD", "GBPUSD"], login=12345)],
    ["M1", "H1"],
    count=500,
    retry_count=3,
)
eurusd_m1 = rates["EURUSD", "M1"]  # closed bars only

• Credential resolution: use resolve_account_spec() / resolve_account_specs() to merge explicit override values over AccountSpec fields and expand ${ENV_VAR} placeholders (via substitute_env_placeholders()), raising ValueError for missing variables. This keeps secrets out of plan/config files without coupling to any strategy code. For config dicts or nested structures loaded from YAML/TOML, use substitute_mapping_values(data, keys={"login", "password"}) to expand placeholders only for caller-specified keys — key names are never hard-coded in mt5cli.
• Throttled history updates: use ThrottledHistoryUpdater to wrap update_history() with a minimum interval_seconds between successful runs (monotonic clock). Call should_update() / update(client, symbols) from an application loop; errors propagate by default, or pass suppress_errors=True to swallow recoverable Mt5*Error, sqlite3.Error, ValueError, and OSError without advancing the throttle (AttributeError / TypeError always propagate, since the client passed to update() must implement the canonical HistoryClient method names). Pass update_backend to inject a custom history update callable (same keyword arguments as update_history) instead of monkey-patching mt5cli.history.update_history.
• Trading helpers: use mt5_session() for both market data and generic execution helpers. It owns initialization and shutdown for sessions it creates; caller-supplied clients remain caller-owned.
• Granularity-keyed rate loading: load_rate_series_by_granularity() builds targets with build_rate_targets(), loads them with load_rate_series_from_sqlite(), and returns a mapping keyed by (symbol | None, granularity_name) such as ("EURUSD", "M1") to reduce downstream boilerplate.
• MT5 session helper: use the mt5_session() context manager to attach to (or, when Mt5Config.path is set, launch) an MT5 terminal, log in, and yield a connected MT5Client that shuts down on exit.
• SQLite export helpers: use export_dataframe_to_sqlite() for append mode, optional index export, and post-write deduplication by key columns.
• Recent ticks and margins: recent_ticks() and minimum_margins() SDK helpers (and matching CLI commands) cover common downstream read-only queries.

Requirements

• Python 3.11+
• Windows OS (MetaTrader 5 requirement)
• MetaTrader 5 platform installed

Migration note for downstream trading apps

Replace local MT5 lifecycle and trading helper code with mt5cli imports:

# Before (local application helpers)
# with local_mt5_session(config) as client:
#     side = local_detect_position_side(client, symbol)
#     sizing = local_calculate_margin_and_volume(client, symbol, unit_ratio, preserved_ratio)
#     limits = local_determine_order_limits(client, symbol, side, sl_ratio, tp_ratio)

# After (mt5cli shared layer)
from pdmt5 import Mt5Config
from mt5cli import (
    calculate_margin_and_volume,
    detect_position_side,
    determine_order_limits,
    mt5_session,
)

with mt5_session(Mt5Config(path=terminal_path, login=login)) as client:
    side = detect_position_side(client, symbol)
    sizing = calculate_margin_and_volume(
        client, symbol, unit_margin_ratio=0.5, preserved_margin_ratio=0.2
    )
    if side is not None:
        limits = determine_order_limits(
            client,
            symbol,
            side,
            stop_loss_limit_ratio=0.01,
            take_profit_limit_ratio=0.02,
        )

Throttled history updates use a separate read-only session:

from pdmt5 import Mt5Config

from mt5cli import ThrottledHistoryUpdater, mt5_session

updater = ThrottledHistoryUpdater(
    output="history.db", interval_seconds=60, suppress_errors=True
)
with mt5_session(Mt5Config(login=login)) as client:
    updater.update(client, ["EURUSD"])

Read-only collectors can keep using mt5_session() and MT5Client.

Development

git clone https://github.com/dceoy/mt5cli.git
cd mt5cli
uv sync

License

MIT