feat(toolkit): ship macro_lactone_toolkit package

Unify macrolactone detection, numbering, fragmentation, and
splicing under the installable macro_lactone_toolkit package.

- replace legacy src.* modules with the new package layout
- add analyze/number/fragment CLI entrypoints and pixi tasks
- migrate tests, README, and scripts to the new package API

tests/test_cli.py

@@ -0,0 +1,74 @@
+from __future__ import annotations
+
+import json
+import subprocess
+import sys
+
+import pandas as pd
+
+from .helpers import build_ambiguous_smiles, build_macrolactone
+
+
+def run_cli(*args: str) -> subprocess.CompletedProcess[str]:
+    return subprocess.run(
+        [sys.executable, "-m", "macro_lactone_toolkit.cli", *args],
+        capture_output=True,
+        text=True,
+        check=False,
+    )
+
+
+def test_cli_smoke_commands():
+    built = build_macrolactone(16, {5: "methyl"})
+
+    analyze = run_cli("analyze", "--smiles", built.smiles)
+    assert analyze.returncode == 0, analyze.stderr
+    analyze_payload = json.loads(analyze.stdout)
+    assert analyze_payload["valid_ring_sizes"] == [16]
+
+    number = run_cli("number", "--smiles", built.smiles)
+    assert number.returncode == 0, number.stderr
+    number_payload = json.loads(number.stdout)
+    assert number_payload["ring_size"] == 16
+    assert number_payload["position_to_atom"]["1"] >= 0
+
+    fragment = run_cli("fragment", "--smiles", built.smiles, "--parent-id", "cli_1")
+    assert fragment.returncode == 0, fragment.stderr
+    fragment_payload = json.loads(fragment.stdout)
+    assert fragment_payload["parent_id"] == "cli_1"
+    assert fragment_payload["ring_size"] == 16
+    assert fragment_payload["fragments"][0]["fragment_smiles_labeled"]
+
+
+def test_cli_fragment_csv_skips_ambiguous_and_records_errors(tmp_path):
+    valid = build_macrolactone(14, {4: "methyl"})
+    ambiguous = build_ambiguous_smiles()
+    input_path = tmp_path / "molecules.csv"
+    output_path = tmp_path / "fragments.csv"
+    errors_path = tmp_path / "errors.csv"
+
+    pd.DataFrame(
+        [
+            {"id": "valid_1", "smiles": valid.smiles},
+            {"id": "ambiguous_1", "smiles": ambiguous},
+        ]
+    ).to_csv(input_path, index=False)
+
+    completed = run_cli(
+        "fragment",
+        "--input",
+        str(input_path),
+        "--output",
+        str(output_path),
+        "--errors-output",
+        str(errors_path),
+    )
+
+    assert completed.returncode == 0, completed.stderr
+
+    fragments = pd.read_csv(output_path)
+    errors = pd.read_csv(errors_path)
+
+    assert set(fragments["parent_id"]) == {"valid_1"}
+    assert errors.loc[0, "parent_id"] == "ambiguous_1"
+    assert errors.loc[0, "error_type"] == "AmbiguousMacrolactoneError"