feat: add strategy and backtest services

- strategy_service.py combines opportunity + portfolio signals into
  unified buy/sell/hold recommendations
- backtest_service.py runs walk-forward backtests on historical datasets
  with virtual cash and positions
- CLI adds `strategy` and `backtest` commands with `--decision-interval`
  and other tuning parameters
- Add tests for both services and CLI dispatch
- Update CLAUDE.md with new architecture docs
- Optimize model weights via opportunity optimizer

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

CLAUDE.md

@@ -4,7 +4,7 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
 
 ## Development commands
 
-- **Install (dev):** `pip install -e ".[dev]"`
+- **Install (dev):** `pip install -e ".[dev]"` or `conda env create -f environment.yml && conda activate coinhunter`
 - **Run CLI locally:** `python -m coinhunter --help`
 - **Run tests:** `pytest` or `python -m pytest tests/`
 - **Run single test file:** `pytest tests/test_cli.py -v`
@@ -16,18 +16,20 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
 
 CoinHunter V2 is a Binance-first crypto trading CLI with a flat, direct architecture:
 
-- **`src/coinhunter/cli.py`** — Single entrypoint (`main()`). Uses `argparse` to parse commands and directly dispatches to service functions. There is no separate `commands/` adapter layer.
+- **`src/coinhunter/cli.py`** — Single entrypoint (`main()`). Uses `argparse` to parse commands and directly dispatches to service functions. There is no separate `commands/` adapter layer. Top-level commands include `init`, `config`, `account`, `market`, `buy`, `sell`, `portfolio`, `opportunity`, `strategy`, `backtest`, `catlog`, `upgrade`, and `completion`.
 - **`src/coinhunter/services/`** — Contains all domain logic:
   - `account_service.py` — balances, positions, overview
   - `market_service.py` — tickers, klines, scan universe, symbol normalization
   - `signal_service.py` — shared market signal scoring used by scan and portfolio analysis
   - `portfolio_service.py` — held-position review and add/hold/trim/exit recommendations
-  - `trade_service.py` — spot and USDT-M futures order execution
+  - `trade_service.py` — spot order execution only
   - `opportunity_service.py` — market scanning and entry/watch/skip recommendations
-- **`src/coinhunter/binance/`** — Thin wrappers around official Binance connectors:
-  - `spot_client.py` wraps `binance.spot.Spot`
-  - `um_futures_client.py` wraps `binance.um_futures.UMFutures`
-  Both normalize request errors into `RuntimeError` and handle single/multi-symbol ticker responses.
+  - `opportunity_dataset_service.py` — historical kline dataset collection for backtesting
+  - `opportunity_evaluation_service.py` — walk-forward evaluation and model-weight optimization
+  - `research_service.py` — external research signal providers for opportunity scoring
+  - `strategy_service.py` — combines opportunity scanning and portfolio analysis into unified buy/sell/hold trade signals
+  - `backtest_service.py` — walk-forward backtest engine using historical kline datasets with virtual cash and positions
+- **`src/coinhunter/binance/spot_client.py`** — Thin wrapper around `binance.spot.Spot`. Normalizes request errors into `RuntimeError` and handles single/multi-symbol ticker responses.
 - **`src/coinhunter/config.py`** — `load_config()`, `get_binance_credentials()`, `ensure_init_files()`.
 - **`src/coinhunter/runtime.py`** — `RuntimePaths`, `get_runtime_paths()`, `print_json()`.
 - **`src/coinhunter/audit.py`** — Writes JSONL audit events to dated files.
@@ -39,6 +41,7 @@ User data lives in `~/.coinhunter/` by default (override with `COINHUNTER_HOME`)
 - `config.toml` — runtime, binance, trading, signal, opportunity, and portfolio settings
 - `.env` — `BINANCE_API_KEY` and `BINANCE_API_SECRET`
 - `logs/audit_YYYYMMDD.jsonl` — structured audit log
+- `logs/dry-run/audit_YYYYMMDD.jsonl` — dry-run audit log
 
 Run `coinhunter init` to generate the config and env templates.
 
@@ -46,8 +49,10 @@ Run `coinhunter init` to generate the config and env templates.
 
 - **Symbol normalization:** `market_service.normalize_symbol()` strips `/`, `-`, `_`, and uppercases the symbol. CLI inputs like `ETH/USDT`, `eth-usdt`, and `ETHUSDT` are all normalized to `ETHUSDT`.
 - **Dry-run behavior:** Trade commands support `--dry-run`. If omitted, the default falls back to `trading.dry_run_default` in `config.toml`.
-- **Client injection:** Service functions accept `spot_client` / `futures_client` as keyword arguments. This enables easy unit testing with mocks.
-- **Error handling:** Binance client wrappers catch `requests.exceptions.SSLError` and `RequestException` and re-raise as human-readable `RuntimeError`. The CLI catches all exceptions in `main()` and prints `error: {message}` to stderr with exit code 1.
+- **Client injection:** Service functions accept `spot_client` as a keyword argument. This enables easy unit testing with mocks.
+- **Error handling:** `spot_client.py` catches `requests.exceptions.SSLError` and `RequestException` and re-raises as human-readable `RuntimeError`. The CLI catches all exceptions in `main()` and prints `error: {message}` to stderr with exit code 1.
+- **Ticker API fallback:** `spot_client.ticker_stats()` uses `rolling_window_ticker` for symbol-specific queries and `ticker_24hr` for full-market scans (no symbols).
+- **Output modes:** All commands support `--agent` for JSON output and `--doc` to print the command's output schema.
 
 ## Testing
 
@@ -56,8 +61,12 @@ Tests live in `tests/` and use `unittest.TestCase` with `unittest.mock.patch`. T
 - `test_cli.py` — parser smoke tests and dispatch behavior
 - `test_config_runtime.py` — config loading, env parsing, path resolution
 - `test_account_market_services.py` — balance/position/ticker/klines logic with mocked clients
-- `test_trade_service.py` — spot and futures trade execution paths
+- `test_trade_service.py` — spot trade execution paths
 - `test_opportunity_service.py` — portfolio and scan scoring logic
+- `test_opportunity_dataset_service.py` — dataset collection and walk-forward evaluation
+- `test_opportunity_evaluation_service.py` — model weight optimization
+- `test_strategy_service.py` — combined signal generation logic
+- `test_backtest_service.py` — historical backtest engine
 
 ## Notes