{
  "markdown": "# TradeMemory Plugin\n\nPersistent memory + autonomous strategy evolution for AI traders. 200+ trading MCP servers execute. None remember. TradeMemory does.\n\n## Installation\n\n### From GitHub (recommended)\n\n```bash\ngit clone https://github.com/mnemox-ai/tradememory-plugin.git\nclaude --plugin-dir ./tradememory-plugin\n```\n\n### Manual\n\nCopy the plugin directory into your project or pass it directly:\n\n```bash\nclaude --plugin-dir /path/to/tradememory-plugin\n```\n\n### MCP Server (standalone, no plugin needed)\n\n```bash\npip install tradememory-protocol\nclaude mcp add tradememory -- uvx tradememory-protocol\n```\n\n## Commands\n\n| Command | Description |\n|---------|-------------|\n| `/record-trade [details]` | Record a completed trade into all 5 OWM memory layers |\n| `/recall [context]` | Recall similar past trades, ranked by outcome-weighted score |\n| `/performance [strategy]` | Generate strategy performance report with behavioral analysis |\n| `/evolve [symbol] [tf] [gens]` | Discover new trading strategies from raw OHLCV data |\n| `/daily-review [date]` | AI-powered daily reflection on trades and behavioral patterns |\n\n## Skills\n\n### Trading Memory\n| Skill | Description |\n|-------|-------------|\n| **trading-memory** | OWM architecture, 5 memory types, recall scoring, behavioral baselines |\n| **evolution-engine** | LLM-powered strategy discovery, vectorized backtesting, OOS validation |\n| **risk-management** | Affective state monitoring, tilt detection, position sizing, behavioral guardrails |\n\n## MCP Tools (15 total)\n\n### Core Memory (4)\n- `store_trade_memory` — Store a trade with context\n- `recall_similar_trades` — Find past trades matching current context\n- `get_strategy_performance` — Aggregate stats per strategy\n- `get_trade_reflection` — Deep-dive into a trade's reasoning\n\n### OWM Cognitive Memory (6)\n- `remember_trade` — Store across all 5 OWM memory layers\n- `recall_memories` — Outcome-weighted recall\n- `get_behavioral_analysis` — Disposition ratio, hold times, Kelly criterion\n- `get_agent_state` — Confidence, drawdown, streaks, risk appetite\n- `create_trading_plan` — Prospective trading plans\n- `check_active_plans` — Evaluate plans against current conditions\n\n### Evolution Engine (5)\n- `evolution_fetch_market_data` — Fetch OHLCV from Binance\n- `evolution_discover_patterns` — LLM-powered pattern discovery\n- `evolution_run_backtest` — Vectorized backtesting\n- `evolution_evolve_strategy` — Full evolution loop\n- `evolution_get_log` — Evolution history and graveyard\n\n## Example Workflows\n\n### Record and Learn\n```\n/record-trade XAUUSD long 5180 5210 +$150\n\n# Stores trade, updates all memory layers, shows similar past trades\n```\n\n### Pre-Trade Check\n```\n/recall London session breakout, high volatility, XAUUSD trending up\n\n# Returns past trades in similar conditions, ranked by P&L outcome\n```\n\n### Strategy Evolution\n```\n/evolve BTCUSDT 1h 3\n\n# Discovers patterns → backtests → selects → mutates × 3 generations\n# Validates out-of-sample → graduates survivors\n```\n\n### End of Day\n```\n/daily-review today\n\n# Analyzes today's trades, checks behavioral drift, updates affective state\n```\n\n## Requirements\n\n- Python 3.10+\n- `pip install tradememory-protocol`\n- Optional: `ANTHROPIC_API_KEY` for LLM reflections and Evolution Engine\n\n## Links\n\n- [Plugin repo](https://github.com/mnemox-ai/tradememory-plugin)\n- [Core protocol](https://github.com/mnemox-ai/tradememory-protocol)\n- [PyPI](https://pypi.org/project/tradememory-protocol/)\n- [Tutorial](https://github.com/mnemox-ai/tradememory-protocol/blob/master/docs/TUTORIAL.md)\n",
  "bytes": 3557,
  "sha": "4019fe8cc97c88bbb14ba072fef162cedac09388ceb1493150c6d7305fcb3fb5",
  "repo_slug": "mnemox-ai/tradememory-plugin",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_mnemox_ai_tradememory_plugin_tradememory_8cd51d41/readme"
}