Build trading algorithms from natural language
Describe momentum, mean-reversion, DCA, breakout, or custom market logic and generate inspectable Python strategy code.
v0.6.7 — now available on npm
Ship trading algorithms from your terminal
An AI agent that writes, validates, and deploys strategies. You bring the idea and the API key.
5 brokerages pre-integrated · AST-based lookahead detection · BYOK
jaiminpatel — Finny — .opencode — node — 203×59finny
FINNY
trading agent
⌘Home
AMy Algos
BBacktest
PPortfolio
SSessions
UUpgrade
FFeedback
SSettings
Think Strategies, Ship Alpha
v0.6.7 · AI trading agent
tab agentsctrl+p commands
Recent algorithms
hype-breakout-pullback
v1 · draft
usd-trend-follower
v1 · draft
eth-trend-pullback
v1 · draft
geopolitical-crisis-volatility
v1 · draft
Portfolio
eth-trend-pullback
+1.43%ROI · paper · 10,080mins
Click to view live session
v 0.6.7
Supported Brokerages
Polymarket
Alpaca
Binance
Questrade
Interactive Brokers
AI trading assistant
Claude Code-style workflow for financial markets.
Finny is for traders, developers, and market builders who want to move from strategy idea to generated code, validation, and backtesting without stitching together prompts, notebooks, and scripts manually.
Backtest before you trust a strategy
Use historical data and metrics like Sharpe ratio, max drawdown, win rate, and profit factor to evaluate trading ideas.
Keep control of models, keys, and execution
Bring your own AI provider keys, review generated code, and connect broker workflows deliberately instead of using a black box.
How Finny Works
Describe what you want in plain English. Finny generates, validates, and saves your trading strategy for automated financial analysis and backtesting.
3 Agent Modes
Build, Research, or Chat — press Tab to switch. Build generates immediately. Research analyzes first. Chat answers questions.
Multi-Broker Support
Polymarket and Alpaca live today. Questrade and Interactive Brokers coming next. Trade across markets from one interface.
Auto-Validation
Every strategy is checked for syntax errors, forbidden imports, dangerous calls, and trading pitfalls like lookahead bias.
Backtesting Engine
Historical data via yfinance. Metrics include Sharpe Ratio, Max Drawdown, Win Rate, and Profit Factor.
Cloud Storage
Algorithms saved to Convex with version tracking. Compare iterations, roll back changes, manage your strategy library.
BYOK
Bring your own API keys. Works with Anthropic, OpenAI, Google, or local models. Your keys, your choice.
Commands
Built-in slash commands for managing your strategies and running backtests.
/algosList saved algorithms
/codeView algorithm source code
/backtestRun historical backtest
/reviewReview strategy changes
/initInitialize project rules
Strategy Templates
7 built-in templates to jumpstart your algorithm. Or go fully custom.
MomentumRSI momentum — buys oversold, sells overbought
Mean ReversionBollinger Bands — buys lower band, sells upper
BreakoutDonchian channel — buys new highs, sells new lows
DCADollar-cost averaging with profit-target exit
Golden CrossSMA 50/200 crossover signals
ScalpingEMA scalping with tight stops
CustomMinimal skeleton — implement your own logic
Start Building Trading Algorithms in Seconds
Install Finny, describe your strategy, and let the AI trading assistant generate a backtestable workflow. Crypto and stocks supported out of the box.
Run finny in your terminal after installation to launch the TUI.
Try: "Follow Trump's news feed and build a bot that trades on his announcements."