agiprolabs/trading-visualization
Professional trading charts including candlesticks, equity curves, drawdowns, correlation heatmaps, and return distributions
npx skills add https://github.com/agiprolabs/claude-trading-skills --skill trading-visualizationVisualization is the primary interface between a trader and their data. Charts reveal patterns that tables and numbers cannot: breakdowns in strategy, regime transitions, clustering of losses, and the shape of risk. A well-designed chart communicates more in a glance than a page of statistics.
Three uses of trading charts:
| Chart Type | Purpose | Library |
|------------|---------|---------|
| Candlestick | OHLCV price action with overlays | mplfinance |
| Equity curve | Portfolio value over time | matplotlib |
| Drawdown | Underwater equity plot | matplotlib |
| Return distribution | Histogram + normal fit | matplotlib |
| Correlation heatmap | Cross-asset correlation matrix | matplotlib / seaborn |
| Trade markers | Entry/exit points on price chart | mplfinance / matplotlib |
| Indicator panels | RSI, MACD below price chart | mplfinance |
| Position timeline | When positions were held | matplotlib |
Best for candlestick charts. Built on matplotlib with finance-specific defaults.
uv pip install mplfinance
import mplfinance as mpf
# Basic candlestick from a DataFrame with DatetimeIndex
# Columns: Open, High, Low, Close, Volume
mpf.plot(df, type="candle", volume=True, style="charles")
Key features:
addplot for overlays (moving averages, Bollinger Bands)mpf.make_mpf_style()General purpose, most flexible. Use when you need full control over layout.
uv pip install matplotlib
import matplotlib.pyplot as plt
fig, axes = plt.subplots(2, 1, figsize=(14, 8), height_ratios=[3, 1],
sharex=True)
axes[0].plot(dates, equity, color="#00ff88")
axes[1].fill_between(dates, drawdown, 0, color="#ff4444", alpha=0.5)
Interactive charts rendered as HTML. Best for exploration and dashboards.
uv pip install plotly
import plotly.graph_objects as go
fig = go.Figure(data=[go.Candlestick(
x=df.index, open=df["Open"], high=df["High"],
low=df["Low"], close=df["Close"]
)])
fig.update_layout(template="plotly_dark")
fig.write_html("chart.html")
Trading terminals use dark backgrounds by default. All charts in this skill follow that convention.
import matplotlib.pyplot as plt
plt.style.use("dark_background")
plt.rcParams.update({
"figure.facecolor": "#1a1a2e",
"axes.facecolor": "#1a1a2e",
"axes.edgecolor": "#333333",
"grid.color": "#333333",
"grid.alpha": 0.4,
"text.color": "#e0e0e0",
"xtick.color": "#aaaaaa",
"ytick.color": "#aaaaaa",
})
| Element | Color | Hex |
|---------|-------|-----|
| Bullish / profit | Green | #00ff88 |
| Bearish / loss | Red | #ff4444 |
| Neutral / info | Blue | #4488ff |
| Warning | Amber | #ffaa00 |
| MA short | Orange | #ff6600 |
| MA long | Blue | #3399ff |
| MA signal | Yellow | #ffcc00 |
See references/styling_guide.md for complete typography, layout ratios, and export settings.
Most trading charts need multiple synchronized panels — price on top, volume in the middle, indicators at the bottom.
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
fig = plt.figure(figsize=(14, 10))
gs = gridspec.GridSpec(3, 1, height_ratios=[3, 1, 1], hspace=0.05)
ax_price = fig.add_subplot(gs[0])
ax_volume = fig.add_subplot(gs[1], sharex=ax_price)
ax_rsi = fig.add_subplot(gs[2], sharex=ax_price)
# Hide x-tick labels on upper panels
ax_price.tick_params(labelbottom=False)
ax_volume.tick_params(labelbottom=False)
| Layout | Ratios | Use Case |
|--------|--------|----------|
| Price + Volume | [3, 1] | Simple OHLCV chart |
| Price + Volume + Indicator | [3, 1, 1] | Standard analysis view |
| Equity + Drawdown | [2, 1] | Performance review |
| Price + RSI + MACD | [3, 1, 1] | Full indicator stack |
import mplfinance as mpf
import pandas as pd
# df: DataFrame with DatetimeIndex, columns Open/High/Low/Close/Volume
ema20 = df["Close"].ewm(span=20).mean()
ema50 = df["Close"].ewm(span=50).mean()
ap = [
mpf.make_addplot(ema20, color="#ff6600", width=1.2),
mpf.make_addplot(ema50, color="#3399ff", width=1.2),
]
style = mpf.make_mpf_style(
base_mpf_style="nightclouds",
marketcolors=mpf.make_marketcolors(
up="#00ff88", down="#ff4444",
wick={"up": "#00ff88", "down": "#ff4444"},
edge={"up": "#00ff88", "down": "#ff4444"},
volume={"up": "#00ff88", "down": "#ff4444"},
),
facecolor="#1a1a2e", figcolor="#1a1a2e",
gridcolor="#333333", gridstyle="--",
)
mpf.plot(df, type="candle", style=style, addplot=ap,
volume=True, figsize=(14, 8),
title="Token / SOL — 15m", savefig="candles.png")
import numpy as np
import matplotlib.pyplot as plt
def plot_equity_drawdown(equity: pd.Series, title: str = "Portfolio") -> plt.Figure:
"""Plot equity curve with drawdown panel below."""
peak = equity.cummax()
drawdown = (equity - peak) / peak
fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(14, 8),
height_ratios=[2, 1], sharex=True)
ax1.plot(equity.index, equity, color="#00ff88", linewidth=1.5)
ax1.plot(equity.index, peak, color="#555555", linewidth=0.8,
linestyle="--", label="Peak")
ax1.set_title(title, fontsize=14, fontweight="bold", color="white")
ax1.set_ylabel("Portfolio Value", fontsize=11)
ax1.legend(loc="upper left")
ax1.grid(True, alpha=0.3)
ax2.fill_between(equity.index, drawdown, 0, color="#ff4444", alpha=0.5)
ax2.set_ylabel("Drawdown", fontsize=11)
ax2.set_xlabel("Date", fontsize=11)
ax2.grid(True, alpha=0.3)
fig.tight_layout()
return fig
from scipy import stats
def plot_return_distribution(returns: pd.Series) -> plt.Figure:
"""Histogram of returns with normal fit and risk metrics."""
fig, ax = plt.subplots(figsize=(10, 6))
ax.hist(returns, bins=50, density=True, alpha=0.7,
color="#4488ff", edgecolor="#333333")
# Normal fit overlay
mu, sigma = returns.mean(), returns.std()
x = np.linspace(returns.min(), returns.max(), 200)
ax.plot(x, stats.norm.pdf(x, mu, sigma), color="#ffaa00",
linewidth=2, label=f"Normal(μ={mu:.4f}, σ={sigma:.4f})")
# VaR line
var_95 = returns.quantile(0.05)
ax.axvline(var_95, color="#ff4444", linestyle="--",
label=f"VaR 95%: {var_95:.4f}")
ax.set_title("Return Distribution", fontsize=14, fontweight="bold")
ax.set_xlabel("Return", fontsize=11)
ax.legend()
ax.grid(True, alpha=0.3)
fig.tight_layout()
return fig
def plot_correlation_heatmap(returns_df: pd.DataFrame) -> plt.Figure:
"""Correlation matrix heatmap with annotations."""
corr = returns_df.corr()
fig, ax = plt.subplots(figsize=(10, 8))
im = ax.imshow(corr, cmap="RdYlGn", vmin=-1, vmax=1, aspect="auto")
ax.set_xticks(range(len(corr.columns)))
ax.set_yticks(range(len(corr.columns)))
ax.set_xticklabels(corr.columns, rotation=45, ha="right")
ax.set_yticklabels(corr.columns)
for i in range(len(corr)):
for j in range(len(corr)):
ax.text(j, i, f"{corr.iloc[i, j]:.2f}",
ha="center", va="center", fontsize=9,
color="black" if abs(corr.iloc[i, j]) < 0.5 else "white")
fig.colorbar(im, ax=ax, shrink=0.8)
ax.set_title("Correlation Matrix", fontsize=14, fontweight="bold")
fig.tight_layout()
return fig
def plot_trades_on_price(
price: pd.Series,
entries: pd.DataFrame, # columns: date, price, side
exits: pd.DataFrame, # columns: date, price, pnl
) -> plt.Figure:
"""Price chart with entry/exit markers."""
fig, ax = plt.subplots(figsize=(14, 7))
ax.plot(price.index, price, color="#aaaaaa", linewidth=1)
# Entry markers
buy_mask = entries["side"] == "long"
ax.scatter(entries.loc[buy_mask, "date"], entries.loc[buy_mask, "price"],
marker="^", color="#00ff88", s=100, zorder=5, label="Buy")
ax.scatter(entries.loc[~buy_mask, "date"], entries.loc[~buy_mask, "price"],
marker="v", color="#ff4444", s=100, zorder=5, label="Short")
# Exit markers
win_mask = exits["pnl"] > 0
ax.scatter(exits.loc[win_mask, "date"], exits.loc[win_mask, "price"],
marker="x", color="#00ff88", s=80, zorder=5)
ax.scatter(exits.loc[~win_mask, "date"], exits.loc[~win_mask, "price"],
marker="x", color="#ff4444", s=80, zorder=5)
ax.set_title("Trades on Price", fontsize=14, fontweight="bold")
ax.legend()
ax.grid(True, alpha=0.3)
fig.tight_layout()
return fig
| Format | Method | Use Case |
|--------|--------|----------|
| PNG | fig.savefig("chart.png", dpi=150) | Sharing, embedding |
| SVG | fig.savefig("chart.svg") | Editing, scaling |
| HTML | fig.write_html("chart.html") (plotly) | Interactive exploration |
| Inline | plt.show() | Jupyter notebooks |
fig.savefig("chart.png", dpi=150, facecolor=fig.get_facecolor(),
edgecolor="none", bbox_inches="tight")
| Skill | Integration |
|-------|-------------|
| pandas-ta | Compute indicators, pass to addplot overlays |
| vectorbt | Extract equity curve and trade list for visualization |
| portfolio-analytics | Plot Sharpe, drawdown, and return metrics |
| risk-management | Visualize position limits and exposure over time |
| position-sizing | Chart position size vs account equity over time |
| regime-detection | Color background by detected market regime |
| correlation-analysis | Generate correlation heatmaps from return data |
references/chart_recipes.md — Complete code recipes for six common chart typesreferences/styling_guide.md — Dark theme setup, colors, typography, layout, and export settingsscripts/chart_generator.py — Generate four chart types from synthetic data (candlestick, equity, returns, trades)scripts/performance_report.py — Multi-chart performance report with summary statisticsTake agiprolabs/trading-visualization from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.
The instructions reference pip, uv.
Without those the skill loads but fails at the first command.