openfinclaw-cli

English | 中文

Your quant research team, in one prompt.

Research · strategy · backtest · paper trade — ship a complete quant workflow from a single natural-language prompt, inside Claude Code, Cursor, and 20+ AI agents.

🚀 Try it in 60 seconds — zero install

Run a full research → strategy → backtest loop in your browser. No install, no API key, real market data.

Quick Start · Example Prompts · Community · Platforms · vs. other tools

What you get

_{Live output from openfinclaw deepagent research — one prompt: research → strategy → backtest → metrics.}

Example Prompts

Copy-paste any of these into openfinclaw deepagent research "…" (or drop them straight into your AI agent). Each one runs the full research → strategy → backtest loop.

📈 Technical analysis

• Find RSI divergence signals on NVDA in the last 6 months, then backtest them.
• Compare a Bollinger Bands strategy on TSLA vs AAPL over 1 year — which wins?
• Screen the S&P 500 for golden-cross signals this month.

📊 Fundamentals & macro

• Pull Apple's last 8 quarters of revenue, margins, and guidance. Summarize the trend.
• What's driving the NVDA move this quarter — earnings, guidance, or narrative?
• Compare AMD / INTC / NVDA on growth, margin, and valuation.

🎯 Strategy generation

• Design a momentum strategy on US mega-cap tech. Backtest 2y. Tell me where it breaks.
• Write a mean-reversion strategy on BTC and show drawdown behavior through 2022.
• A-shares 沪深 300 日内轮动策略，年化目标 15%，最大回撤 < 10%。

🧪 Backtest & stress-test

• Backtest a 50/200 SMA crossover on SPY from 2015. Include costs and slippage.
• Stress-test my forked strategy against the 2020 and 2022 crashes.

Want a ready-made one? Run openfinclaw leaderboard to browse the community's highest-ranked strategies, then fork any of them.

Quick Start

💡 Want to see it in action before installing? Try DeepAgent in your browser first.

60-second onboarding

npx @openfinclaw/cli@latest install               # wizard + MCP configs + Skill registration + doctor
openfinclaw deepagent +research "盘点 BTC 周线"     # streaming research / strategy / backtest

install is the new one-liner: it runs the interactive wizard, writes MCP configs to every detected AI agent, persists your fch_ key to ~/.openfinclaw/config.json (chmod 600 on Unix), registers OpenFinClaw as a global AI Skill so Claude Code / Cursor auto-trigger it on keywords like quant/backtest/量化, and finishes with a connectivity check.

Note on terminology. The skill_* MCP tools and the openfinclaw skill-install command refer to two different things. The MCP tools manage Hub strategy packages (FEP v2.0 ZIPs published on https://hub.openfinclaw.ai). skill-install registers OpenFinClaw as an AI Agent skill (a SKILL.md file under ~/.claude/skills/). Both live in the codebase under the same word — sorry.

Non-interactive / CI

npx @openfinclaw/cli@latest install --yes \
  --platforms cursor,claude-code \
  --tool-groups deepagent,strategy \
  --api-key fch_xxx \
  --register-skill

Just-the-wizard mode (no Skill registration, no doctor)

npx @openfinclaw/cli init

The wizard:

• Pre-selects platforms when either common install markers match (app bundles, user data dirs, CLI on PATH) or the expected MCP config path already exists.
• Asks once for the unified fch_ key.
• Lets you pick tool groups (deepagent, strategy).

CLI vs MCP: agent platforms load the API key from their MCP env block — that does not modify your shell profile. Resolution order for openfinclaw / serve is: --api-key flag → OPENFINCLAW_API_KEY env var → ~/.openfinclaw/config.json.

2. Manual Configuration

Add to your agent platform's MCP config:

{
  "mcpServers": {
    "openfinclaw": {
      "command": "npx",
      "args": ["@openfinclaw/cli", "serve"],
      "env": {
        "OPENFINCLAW_API_KEY": "fch_your_key_here"
      }
    }
  }
}

A single fch_ key drives both DeepAgent (via Hub Gateway) and the strategy group — no separate DeepAgent key needed.

3. CLI Mode (Human Use)

Step 1 — Install (choose one)

# Option A (recommended): install globally, use the short `openfinclaw` command everywhere
npm install -g @openfinclaw/cli      # or: pnpm add -g @openfinclaw/cli

# Option B: no install — prefix every command with `npx -y @openfinclaw/cli`
#   (slower first run while the package is fetched)

All examples below use the short openfinclaw <cmd> form. If you chose Option B, replace it with npx -y @openfinclaw/cli <cmd>.

Step 2 — Provide your API key (choose one)

# A. Run the init wizard once (writes ~/.openfinclaw/config.json, mode 600)
openfinclaw init

# B. Export for the current shell session
export OPENFINCLAW_API_KEY=fch_your_key_here   # drives both deepagent & strategy

# C. Pass it inline per command
openfinclaw deepagent research "..." --api-key fch_your_key_here

Step 3 — Run commands

# Streaming research / analysis / strategy / backtest — all in one prompt
openfinclaw deepagent research "Research NVDA last 90 days, propose a momentum strategy, backtest 1y, suggest a paper-trade plan"

# Inspect past DeepAgent runs
openfinclaw deepagent backtests
openfinclaw deepagent packages
openfinclaw deepagent download <packageId>

# Service health (public, no key needed)
openfinclaw deepagent health

# Strategy leaderboard (Hub key required)
openfinclaw leaderboard --limit 10

# Diagnose config & connectivity
openfinclaw doctor

# Upgrade to the latest version
openfinclaw update

All CLI commands

The +verb prefix (e.g. deepagent +research) selects the human-friendly streaming path; the bare verbs and the MCP-only research_submit/poll/finalize triplet are kept for scripted/agent use.

Run openfinclaw --help for full usage and options.

4. DeepAgent in depth

DeepAgent shares the same fch_ key as the strategy group — traffic is routed through the Hub Gateway, which validates the key and forwards the request to the DeepAgent backend.

# Save the key (or pass it inline with --api-key)
export OPENFINCLAW_API_KEY=fch_your_key_here

# Streaming research in the terminal (token-by-token)
openfinclaw deepagent research "Write me a Tesla Bollinger Bands strategy and run a backtest"

openfinclaw init writes the unified key to ~/.openfinclaw/config.json. Request a key on the Hub dashboard, or trial the service online at https://hub.openfinclaw.ai/en/chat.

Sample output — one prompt produces strategy definition, backtest metrics, trade-level P&L, and improvement suggestions:

Community: leaderboard → fork → publish

OpenFinClaw ships with a community strategy exchange. Browse what others are running, copy any strategy locally, tweak it, and publish back — think of it as a Hugging Face for quant strategies.

openfinclaw leaderboard --limit 20          # Browse top-ranked strategies
openfinclaw strategy-info <id>              # See how a strategy performs
openfinclaw fork <id>                       # Copy to ./strategies/<slug>
# ... edit strategy.py, tweak fep.yaml ...
openfinclaw validate ./strategies/<slug>    # Pre-flight FEP v2.0 check
openfinclaw publish ./my-strategy.zip       # Ship to the leaderboard
openfinclaw publish-verify --submission-id <id>   # Track backtest progress

Every published strategy is backtested server-side and ranked by live-market-equivalent returns — no self-reported numbers.

Supported Platforms

OpenFinClaw works with any MCP-compatible agent platform:

Platform Config Examples

{
  "mcpServers": {
    "openfinclaw": {
      "command": "npx",
      "args": ["@openfinclaw/cli", "serve", "--tools=deepagent,strategy"],
      "env": {
        "OPENFINCLAW_API_KEY": "fch_xxx"
      }
    }
  }
}

{
  "mcpServers": {
    "openfinclaw": {
      "command": "npx",
      "args": ["@openfinclaw/cli", "serve", "--tools=deepagent,strategy"],
      "env": {
        "OPENFINCLAW_API_KEY": "fch_xxx"
      }
    }
  }
}

{
  "servers": {
    "openfinclaw": {
      "command": "npx",
      "args": ["@openfinclaw/cli", "serve", "--tools=deepagent,strategy"],
      "env": {
        "OPENFINCLAW_API_KEY": "fch_xxx"
      }
    }
  }
}

mcp_servers:
  openfinclaw:
    command: "npx"
    args: ["@openfinclaw/cli", "serve", "--tools=deepagent,strategy"]
    env:
      OPENFINCLAW_API_KEY: "fch_xxx"

Add OpenFinClaw to your MCP config (e.g. ~/.openclaw/mcp.json):

{
  "mcpServers": {
    "openfinclaw": {
      "command": "npx",
      "args": ["@openfinclaw/cli", "serve"],
      "env": {
        "OPENFINCLAW_API_KEY": "fch_xxx"
      }
    }
  }
}

Tool Groups & Context Optimization

Load only the tools you need to minimize token usage:

# DeepAgent only — the one-stop quant agent (~1,400 tokens)
npx @openfinclaw/cli serve --tools=deepagent

# Strategy group only (~1,000 tokens)
npx @openfinclaw/cli serve --tools=strategy

# Multiple groups
npx @openfinclaw/cli serve --tools=deepagent,strategy

# All tools (default)
npx @openfinclaw/cli serve

Architecture

┌─────────────────────────────────┐
│       @openfinclaw/core         │  Pure business logic
│  (zero platform dependencies)   │  DeepAgent client, strategy tools, types
└──────────────┬──────────────────┘
               │
       ┌───────┼───────┐
       ▼       ▼       ▼
   ┌───────┐ ┌─────┐ ┌──────┐
   │  MCP  │ │ CLI │ │ Init │
   │Server │ │Mode │ │Wizard│
   └───┬───┘ └──┬──┘ └──┬───┘
       │        │       │
       ▼        ▼       ▼
   20+ Agent  Terminal  Auto-config
   Platforms   Users    Platforms

The project is a monorepo with two packages:

• @openfinclaw/core — Platform-independent business logic (DeepAgent client, strategy tools, shared types)
• @openfinclaw/cli — MCP Server + CLI + interactive setup wizard

Environment Variables

Get your API key at hub.openfinclaw.ai, or try DeepAgent online first at https://hub.openfinclaw.ai/en/chat.

Development

# Clone and install
git clone https://github.com/mirror29/openfinclaw-cli.git
cd openfinclaw-cli
pnpm install

# Build all packages
pnpm build

# Run CLI locally
OPENFINCLAW_API_KEY=<fch_...> node packages/cli/dist/index.js deepagent health

# Run MCP server locally
OPENFINCLAW_API_KEY=<fch_...> node packages/cli/dist/index.js serve

License

MIT