Prophet Trader

Community

Deprecated AI-driven options trading MCP server (successor: OpenProphet)

How to install ↓ Compare GitHub source repository ↗

Category: Other #201 of 230
Stars: ★ 232 Popular
Runtime: Node.js · Go
Credentials: API key / credential required
License: MIT
Last commit: 7 months ago ⚠
Tools: 32

37FMRS · D

Prophet Trader is a feature-rich but author-deprecated experimental MCP server for AI-driven options trading, bridging Claude Code to a Go backend that executes trades via Alpaca, with Gemini-powered news intelligence and vector-based trade memory. The repo still functions, but future development has moved to OpenProphet, and the author explicitly warns it should be used for paper trading only, with no responsibility taken for real-money losses.

Strongest · Documentation 13/20 Weakest · Security and permissions 3/20

Reliability

8/20

Security and permissions

3/20

Maintenance

4/20

Documentation

13/20

Setup experience

9/20

Why each score

Reliability 8/20: The README describes a call chain between the Node.js MCP server (mcp-server.js) and a Go backend exposing 40+ tools, but the repository shows no verifiable test suite, CI workflow, or documented error-handling behavior. Since this is a static-only review with no committed tests or CI evidence, reliability is capped below 12. The architectural description is internally consistent, but whether declared tool behavior matches actual implementation cannot be confirmed, warranting a middling score.
Security and permissions 3/20: The server stores Alpaca trading credentials and a Gemini API key via .env and exposes tools like place_options_order and place_managed_position that can directly execute real-money trades. The README recommends 'paper trading first,' but the .env template leaves ALPACA_ENDPOINT blank with no enforced paper-trading default, and no order-confirmation step is described anywhere. An autonomous AI agent able to issue potentially irreversible real-money trades without a human confirmation gate is a high-risk, near-red-line pattern (irreversible 'pay' action without confirmation), so the security score is heavily reduced.
Maintenance 4/20: The README opens with an explicit 'THIS IS NOW DEPRECATED' notice, redirecting users to a successor repository (OpenProphet), indicating this repo itself is no longer actively developed and retained only for legacy use. While an MIT license is present, there is no evidence of ongoing maintenance, versioned releases, or a security response channel, so maintenance activity scores low.
Documentation 13/20: The README is substantial, including an architecture diagram, tool reference tables, trading strategy and risk rules, a daily workflow, and example JSON output — good layered documentation. However, it lacks explicit authentication setup detail, a troubleshooting section, and clear rate-limit/cost disclosure (only implicit mention of Gemini API cost), and the repo's own deprecation notice reduces its current applicability, so it falls short of full marks.
Setup experience 9/20: Setup requires building the Go backend (go build), configuring multiple environment variables (Alpaca keys, Gemini key), and relying on .mcp.json to auto-launch the Node MCP server, with steps spread across several components and no concrete .mcp.json example shown. As this is a static-only review with no CI or test evidence that the flow actually works end-to-end, setup ease is capped below 15 and further reduced for multi-step complexity.

Static review · not runListed 2026-08-26

Read the FMRS scoring method →

Fit and risk

What it can accessRuns commands or codeUses the networkConnects to a database

Best for

Developers comfortable with Go and the Alpaca API who want to experiment with AI-driven options strategies in a paper trading sandbox
Engineers interested in an example architecture for wiring Claude Code to a custom trading backend via MCP

Not for

Anyone wanting to trade real money in production — the author explicitly warns to use paper trading only
Users who need active maintenance or updates, since this repo is deprecated and development has moved to OpenProphet
Users without an Alpaca trading account and API keys, or unfamiliar with compiling and running a Go backend

Required permissions

Alpaca API keys (permission to place/cancel orders and read positions and account balance)
Optional Gemini API key for AI-based news cleaning and summarization
Local filesystem write access for the SQLite database and activity_logs/decisive_actions logs
Outbound network access to Alpaca, Google News, MarketWatch, and Gemini services

Risks and side effects

Misconfiguring a live (non-paper) Alpaca endpoint can result in real financial losses; the author provides no warranty
The project is marked deprecated by its author and no longer receives updates or fixes — review the code yourself before relying on it
AI-generated trading decisions may be flawed and should be reviewed before execution
Alpaca and Gemini API keys are stored in plaintext in the .env file and must be protected from exposure

Setup

Before you start

Runtime：Node.js · Go

ALPACA_PUBLIC_KEY requiredsecret Alpaca API public key from your Alpaca account; required and secret

ALPACA_SECRET_KEY requiredsecret Alpaca API secret key from your Alpaca account; required and secret

ALPACA_ENDPOINT required Alpaca API endpoint URL; start with the paper trading endpoint; required

GEMINI_API_KEY optionalsecret Google Gemini API key for AI news cleaning; optional - intelligence tools fall back to raw news without it

Create a .env file in the repo root with ALPACA_PUBLIC_KEY, ALPACA_SECRET_KEY, ALPACA_ENDPOINT (start with the paper trading endpoint), and an optional GEMINI_API_KEY
Build the Go backend with go build -o prophet_bot ./cmd/bot
Run ./prophet_bot to start the trading backend
Claude Code launches mcp-server.js automatically via its .mcp.json configuration (the README doesn't show the exact .mcp.json contents)
In Claude Code, call a tool like get_account to confirm the connection.

Check that it works

After starting Claude Code, confirm tools like get_account and place_options_order appear in the tool list, then call get_account to prove the connection to the Go backend and Alpaca.

Troubleshooting

Verify ALPACA_PUBLIC_KEY / ALPACA_SECRET_KEY / ALPACA_ENDPOINT in .env are set correctly, pointed at the paper trading endpoint during testing
Ensure the Go backend (prophet_bot) is built and running before invoking MCP tools
If GEMINI_API_KEY is unset, intelligence tools return raw (non-AI-cleaned) news — this is expected behavior
Check whether the default port 4534 is already in use by another process
Inspect the activity_logs/ and decisive_actions/ directories for logs when debugging unexpected tool behavior

Things to try

Once connected, you can ask your AI assistant things like:

Check my account status and open options positions
Get the options chain and latest quote for SPY
Place a limit options order with place_options_order and log the decision
Analyze a stock technically and summarize today's market news

Tools 32

place_options_order writes

Buy or sell options contracts.

place_managed_position writes

Open a position with automatic stop-loss/take-profit management.

close_managed_position writes

Close a managed position at market price.

cancel_order writes

Cancel a pending order.

place_buy_order writes

Buy stock (not used in this options-only workflow).

place_sell_order writes

Sell stock (not used in this options-only workflow).

get_account read-only

Retrieve portfolio value, cash, and buying power.

get_options_positions read-only

List all open options positions.

Show 24 more tools

get_options_position read-only

Get details for a single options position.

get_options_chain read-only

Retrieve available option contracts for an underlying symbol.

get_orders read-only

Retrieve order history.

get_quote read-only

Get a real-time stock quote.

get_latest_bar read-only

Get the latest OHLCV price bar.

get_historical_bars read-only

Retrieve historical price data.

get_managed_positions read-only

List all managed positions with their status.

get_quick_market_intelligence read-only

Get an AI-cleaned MarketWatch news summary (fast).

analyze_stocks read-only

Run technical analysis plus news context and recommendations for stocks.

search_news read-only

Search Google News by keyword.

get_cleaned_news read-only

Get aggregated, AI-cleaned news from multiple sources.

get_marketwatch_topstories read-only

Get MarketWatch top stories.

get_marketwatch_realtime read-only

Get real-time MarketWatch headlines.

get_marketwatch_bulletins read-only

Get MarketWatch breaking news bulletins.

get_marketwatch_marketpulse read-only

Get quick MarketWatch market pulse updates.

aggregate_and_summarize_news read-only

Aggregate and summarize news for a custom topic or symbol.

find_similar_setups read-only

Find past trades with setups similar to the current one via vector search.

store_trade_setup writes

Store a completed trade setup for future similarity search.

get_trade_stats read-only

Get win rate and profit factor statistics filtered by criteria.

log_decision writes

Log a trading decision along with its reasoning.

log_activity writes

Log an activity entry to the daily journal.

get_activity_log read-only

Retrieve today's activity log.

wait read-only

Pause execution for up to 300 seconds.

get_datetime read-only

Get the current date/time in the US Eastern timezone.

Use cases

Experimenting with AI-assisted, bidirectional (calls and puts) options strategies in a paper trading account

Combining technical analysis with AI-cleaned news for stock/options research

Using vector search to recall similar historical trade setups when reviewing strategy

Having Claude Code log trade reasoning and generate daily activity journals

Supported clients

Claude Code

Listed from the project's documentation, not tested by this site.

Overview

Prophet Trader is an experimental AI-powered options trading system that connects Claude Code to a Go trading backend via a Node.js MCP server (mcp-server.js), which in turn calls the Alpaca API to execute trades. It supports LEAPS-style long-dated positions, 0-5 DTE intraday scalping, trading both calls and puts, managed positions with automatic stop-loss/take-profit, Gemini-powered news cleaning and market intelligence, and a local vector-embedding store for retrieving similar past trades ('AI memory'). The README explicitly states this repository is now deprecated in favor of JakeNesler/OpenProphet, though the code still works as an MCP server. The author repeatedly warns this is experimental and strongly recommends paper trading only, never real money.

Similar servers

Hevy MCP 77 · B

Manage your Hevy workout data from AI assistants via MCP.

★ 517 · Tools 26 Compare with this →

Firecrawl MCP Server 75 · B

Firecrawl's official MCP server for web search, scraping, and structured extraction for AI agents

★ 7.6k · Tools 11 Compare with this →

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →

Compartment 75 · B

Encrypted, fully offline long-term memory for AI agents

★ 583 · Tools 14 Compare with this →

Compare with other MCP servers

Source revision 2fb225130004 Data synced 2026-10-19 Read the FMRS scoring method