SMC Multi-Strategy Paper Trader

Paper trading monitors for SMC (Smart Money Concepts) strategies. Includes swing (4H BoS+FVG), day (1H BoS+FVG), and 8D/2S coordinated orchestration. Uses AT...

Security Scan

The skill's files and runtime instructions match its stated purpose (Binance public-data paper trading) and request no extra credentials, with a minor note about unused GitHub-API helper functions that are not declared or documented.

Purpose & Capability

Name/description (SMC paper trader) align with included scripts: all three monitors fetch Binance public klines/ticker data, compute indicators, and write local portfolio/lock JSON files. No unrelated cloud credentials or unrelated binaries are required.

Instruction Scope

SKILL.md instructs running node scripts and describes portfolio and orchestrator-lock files; the scripts read/write files under ~/.openclaw/workspace/trading which is coherent with the described local paper-dashboard. Note: scripts perform network calls to Binance public APIs (expected) and create/modify local JSON state (portfolio, orchestrator-lock).

Install Mechanism

No install spec or third‑party package downloads are included; code is pure Node.js using built-in https/fs modules. This is low-risk from an install/download perspective.

Credentials

Skill declares no required env vars or credentials and uses only process.env.HOME to locate workspace files. One minor inconsistency: several scripts include helper functions (apiFetch2, apiFetch2Put) that make authenticated GitHub-style requests when passed a token, but no token env var or mechanism is declared in SKILL.md. This appears unused/optional, but it should be reviewed before enabling any token-based behavior.

Persistence & Privilege

The skill persists state locally (portfolio and orchestrator-lock JSON files) under the user's home workspace. always:false and no modification of other skills or global agent config. Local persistence is expected for a paper-trading monitor.

Assessment

This package appears to do what it claims: local paper-trading using Binance public data and local JSON state. Before running: (1) inspect the three scripts locally (they're included) to confirm no unexpected outbound endpoints beyond Binance (the code only calls Binance and contains optional GitHub helpers), (2) be aware the monitors will read/write files under ~/.openclaw/workspace/trading (backup any existing files in that path), (3) if you see or plan to supply any tokens (e.g., for GitHub), do not store secrets in plaintext; the skill does not declare or document any credential usage, so granting a token could enable remote operations, and (4) consider running initially in a sandbox/container or with a disposable HOME path to observe behavior and network traffic. If you want me to, I can search the full files for any calls to apiFetch2/apiFetch2Put usage sites and confirm whether the GitHub helpers are actually reachable/used.

Paper trading system implementing Smart Money Concepts across three strategy types:

• Day trading (1H timeframe) — paper-monitor-v5.js
• Swing trading (4H + Daily timeframe) — paper-monitor-swing.js
• Coordinated 8D/2S (shared capital, 8 day + 2 swing slots) — paper-monitor-coordinated.js

All monitors share an orchestrator-lock file to prevent conflicting positions across strategies on the same symbol.

Architecture

Binance Public API (Kline, Ticker)
        ↓
paper-monitor-[strategy].js
  ├── Fetch candles (1H/4H/1D)
  ├── Detect Break of Structure (BoS)
  ├── Find Fair Value Gaps (FVG)
  ├── Z-score: Volume, TA, FVG quality
  ├── ATR-based SL / TP / Trailing stop
  ├── Orchestrator lock read/write
  └── portfolio-[strategy].json (P&L tracking)
        ↓
paper-dashboard/index.html (4-curve equity chart)

Strategy Configs

Swing (paper-monitor-swing.js)

• Timeframe: Daily BoS + 4H FVG
• Best backtest config: TA≥5, fvgAge=6, srch=4, SL=2.5×ATR, TP=5×ATR, trail=3×ATR
• Max hold: 72h | Max concurrent: 5 | Position size: 10%
• Source: 65-split backtest grid

Day (paper-monitor-v5.js)

• Timeframe: 1H BoS + FVG + CVD z-score
• CVD filter z≥0.5 — reduces trades but improves WR
• Max hold: 12h | Position size: 10%

Coordinated (paper-monitor-coordinated.js)

• Best split: 8 day slots + 2 swing slots (from 65-combination grid)
• Shared orchestrator-lock prevents duplicate symbol entries
• Capital allocation: proportional to slots (80/20)

Orchestrator Lock

// orchestrator-lock.json
{
  "BTCUSDT": { "strategy": "swing", "entryTime": 1742400000000 },
  "ETHUSDT": { "strategy": "day",   "entryTime": 1742405000000 }
}

Lock is read before entry: if symbol is locked by another strategy, skip. On exit (SL/TP/trail/timeout), lock is released.

Portfolio JSON Structure

{
  "balance": 1000,
  "positions": [],
  "history": [],
  "metrics": { "totalTrades": 0, "wins": 0, "losses": 0, "winRate": 0 }
}

Running

# Swing: every 4h
node scripts/paper-monitor-swing.js

# Day: every 1h
node scripts/paper-monitor-v5.js

# Coordinated: every 1h
node scripts/paper-monitor-coordinated.js

Cron Setup (OpenClaw)

Swing:       30 */4 * * *  (XX:30 UTC every 4h)
Day v5:      2 * * * *     (XX:02 UTC every 1h)
Coordinated: 5 * * * *     (XX:05 UTC every 1h)

Files

• scripts/paper-monitor-swing.js — swing strategy
• scripts/paper-monitor-coordinated.js — 8D/2S coordinated
• scripts/paper-monitor-v5.js — day + CVD strategy
• paper-dashboard/index.html — 4-curve equity dashboard

Notes

• Uses Binance public API only — no credentials required
• Start balance: $1,000 paper
• All symbols configurable via TOKENS array
• ATR calculation: Wilder's smoothing (standard)
• TA z-score: composite of EMA cross, RSI, MACD, Bollinger, MFI, Stoch