refactor: address high-priority debt and publish to PyPI

- Fix TOCTOU race conditions by wrapping read-modify-write cycles
  under single-file locks in execution_state, portfolio_service,
  precheck_state, state_manager, and precheck_service.
- Add missing test coverage (96 tests total):
  - test_review_service.py (15 tests)
  - test_check_api.py (6 tests)
  - test_external_gate.py main branches (+10 tests)
  - test_trade_execution.py new commands (+8 tests)
- Unify all agent-consumed JSON messages to English.
- Config-ize hardcoded values (volume filter, schema_version) via
  get_user_config with sensible defaults.
- Add 1-hour TTL to exchange cache with force_new override.
- Add ruff and mypy to dev dependencies; fix all type errors.
- Add __all__ declarations to 11 service modules.
- Sync README with new commands, config tuning docs, and PyPI badge.
- Publish package as coinhunter==1.0.0 on PyPI with MIT license.
- Deprecate coinhunter-cli==1.0.1 with runtime warning.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

tests/test_state_manager.py

@@ -0,0 +1,99 @@
+"""Tests for state_manager precheck utilities."""
+
+from datetime import timedelta, timezone
+
+from coinhunter.services.state_manager import (
+    clear_run_request_fields,
+    sanitize_state_for_stale_triggers,
+    update_state_after_observation,
+)
+from coinhunter.services.time_utils import utc_iso, utc_now
+
+
+class TestClearRunRequestFields:
+    def test_removes_run_fields(self):
+        state = {"run_requested_at": utc_iso(), "run_request_note": "test"}
+        clear_run_request_fields(state)
+        assert "run_requested_at" not in state
+        assert "run_request_note" not in state
+
+
+class TestSanitizeStateForStaleTriggers:
+    def test_no_changes_when_clean(self):
+        state = {"pending_trigger": False}
+        result = sanitize_state_for_stale_triggers(state)
+        assert result["pending_trigger"] is False
+        assert result["_stale_recovery_notes"] == []
+
+    def test_clears_completed_run_request(self):
+        state = {
+            "pending_trigger": True,
+            "run_requested_at": utc_iso(),
+            "last_deep_analysis_at": utc_iso(),
+        }
+        result = sanitize_state_for_stale_triggers(state)
+        assert result["pending_trigger"] is False
+        assert "run_requested_at" not in result
+        assert any("completed run_requested" in note for note in result["_stale_recovery_notes"])
+
+    def test_clears_stale_run_request(self):
+        old = (utc_now() - timedelta(minutes=60)).replace(tzinfo=timezone.utc).isoformat()
+        state = {
+            "pending_trigger": False,
+            "run_requested_at": old,
+        }
+        result = sanitize_state_for_stale_triggers(state)
+        assert "run_requested_at" not in result
+        assert any("stale run_requested" in note for note in result["_stale_recovery_notes"])
+
+    def test_recovers_stale_pending_trigger(self):
+        old = (utc_now() - timedelta(minutes=60)).replace(tzinfo=timezone.utc).isoformat()
+        state = {
+            "pending_trigger": True,
+            "last_triggered_at": old,
+        }
+        result = sanitize_state_for_stale_triggers(state)
+        assert result["pending_trigger"] is False
+        assert any("stale pending_trigger" in note for note in result["_stale_recovery_notes"])
+
+
+class TestUpdateStateAfterObservation:
+    def test_updates_last_observed_fields(self):
+        state = {}
+        snapshot = {
+            "generated_at": "2024-01-01T00:00:00Z",
+            "snapshot_hash": "abc",
+            "positions_hash": "pos123",
+            "candidates_hash": "can456",
+            "portfolio_value_usdt": 100.0,
+            "market_regime": "neutral",
+            "positions": [],
+            "top_candidates": [],
+        }
+        analysis = {"should_analyze": False, "details": [], "adaptive_profile": {}}
+        result = update_state_after_observation(state, snapshot, analysis)
+        assert result["last_observed_at"] == snapshot["generated_at"]
+        assert result["last_snapshot_hash"] == "abc"
+
+    def test_sets_pending_trigger_when_should_analyze(self):
+        state = {}
+        snapshot = {
+            "generated_at": "2024-01-01T00:00:00Z",
+            "snapshot_hash": "abc",
+            "positions_hash": "pos123",
+            "candidates_hash": "can456",
+            "portfolio_value_usdt": 100.0,
+            "market_regime": "neutral",
+            "positions": [],
+            "top_candidates": [],
+        }
+        analysis = {
+            "should_analyze": True,
+            "details": ["price move"],
+            "adaptive_profile": {},
+            "hard_reasons": ["moon"],
+            "signal_delta": 1.5,
+        }
+        result = update_state_after_observation(state, snapshot, analysis)
+        assert result["pending_trigger"] is True
+        assert result["pending_reasons"] == ["price move"]