Publish matching Windows CLI MCP package with document helper

README.md

@@ -221,9 +221,10 @@ Your agent receives the returned excerpts; cloud agents handle them according
 to that client's configuration.
 
 On Windows x64, your agent can use the optional
-[CLI/MCP download](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.0/searchi-cli_0.1.0_x64.exe)
-with its [own checksum](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.0/CLI-SHA256SUMS.txt).
-No Rust build is needed. Install the desktop app first for the model setup and
+[CLI/MCP download](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.3/searchi-cli_0.1.0_x64-013.zip)
+with its [own checksum](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.3/CLI-SHA256SUMS-0.1.3.txt).
+Extract the ZIP alongside the installed desktop runtime helpers; it includes
+the document extraction helper. No Rust build is needed. Install the desktop app first for the model setup and
 runtime helpers, then connect the CLI to that app's existing index.
 
 ## Optional answers, your choice

docs/AGENT-INTEGRATION.md

@@ -8,10 +8,10 @@ same actual Rust/Google search engine.
 ## Configure an existing index
 
 Windows x64 users can download the optional
-[CLI/MCP executable](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.0/searchi-cli_0.1.0_x64.exe)
-and verify its [separate checksum file](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.0/CLI-SHA256SUMS.txt).
-Keep it alongside the installed desktop's runtime helpers; rename it to
-`searchi-cli.exe` if desired. The desktop installer itself is unchanged.
+[CLI/MCP ZIP](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.3/searchi-cli_0.1.0_x64-013.zip)
+and verify its [separate checksum file](https://tea.karlcom.de/karl/searchi/releases/download/v0.1.3/CLI-SHA256SUMS-0.1.3.txt).
+Extract it alongside the installed desktop's runtime helpers. Use
+`searchi-cli.exe`; keep `searchi-extract-worker.exe` beside it for PDF/DOCX indexing.
 Use the desktop to finish model setup and semantic indexing before connecting.
 Pass that existing `searchi.sqlite`, model cache and `searchi-gemma-worker.exe`
 explicitly; avoid creating a second database for the same folders.
@@ -193,3 +193,16 @@ Default tests use genuine SQLite retrieval without downloaded model assets.
 Hybrid mode uses the same real Google service; it has no simulated embeddings.
 Native Linux/macOS MCP execution and cross-platform packaging require their own
 recorded checks; Windows CPU checks do not establish those releases.
+
+## Current Windows CLI package
+
+The optional ZIP published alongside 0.1.3 contains CLI component 0.1.0 and
+its verified document helper. Locked release build passed. Actual extracted
+CLI indexed PDF/DOCX without helper overrides or extraction failures; MCP
+initialization, three read-only tools, document search and cited source reads
+passed on CPU. Two packaging tests passed normally and with Python `-O`.
+Archive: 8784871 bytes, SHA-256
+`4ae415d6cf7d3c9e566933e1de84a85c08c1a92b3f89b7f285384795bab30d64`.
+Public full-byte archive and checksum downloads matched. This does not prove
+clean-machine installation or GPU behavior. Earlier evidence describes older
+assets. Rebuild with `python tools/package_windows_cli.py --executable target/release/searchi.exe`.

tools/package_windows_cli.py

@@ -0,0 +1,69 @@
+"""Package the optional Windows CLI/MCP client; reuse the installed desktop models."""
+import argparse
+import hashlib
+import json
+import os
+from pathlib import Path
+import tempfile
+import zipfile
+
+from package_windows import pe_machine
+from stage_extraction_worker import verify
+
+ROOT = Path(__file__).resolve().parents[1]
+MAX_FILE = 64 * 1024 * 1024
+
+
+def package(root, executable, resources, output):
+    root, executable, resources, output = map(Path, (root, executable, resources, output))
+    def read(path):
+        if path.is_symlink() or not path.is_file() or path.absolute() != path.resolve():
+            raise ValueError('Package input missing or linked')
+        with path.open('rb') as stream:
+            data = stream.read(MAX_FILE + 1)
+        if not data or len(data) > MAX_FILE:
+            raise ValueError('Package input empty or oversized')
+        return data
+    cli = read(executable)
+    architecture = {0x8664: 'x86_64', 0xAA64: 'aarch64'}.get(pe_machine(executable))
+    if architecture is None:
+        raise ValueError('Unsupported Windows CLI architecture')
+    expected = verify(resources, root, 'windows', architecture)
+    selected = {'searchi-cli.exe': cli, 'LICENSE.txt': read(root / 'LICENSE')}
+    for name, record in expected.items():
+        data = read(resources / name)
+        if len(data) != record['bytes'] or hashlib.sha256(data).hexdigest() != record['sha256']:
+            raise ValueError('Extraction resource changed while packaging')
+        selected[name] = data
+    # The captured CLI must still match the header used for architecture validation.
+    if read(executable) != cli:
+        raise ValueError('CLI changed while packaging')
+    records = [{'path': name, 'bytes': len(data), 'sha256': hashlib.sha256(data).hexdigest()}
+               for name, data in selected.items()]
+    output.parent.mkdir(parents=True, exist_ok=True)
+    with tempfile.NamedTemporaryFile(dir=output.parent, suffix='.partial', delete=False) as temporary:
+        temporary_path = Path(temporary.name)
+    try:
+        with zipfile.ZipFile(temporary_path, 'w', zipfile.ZIP_DEFLATED) as archive:
+            for name, data in selected.items():
+                archive.writestr(name, data)
+            archive.writestr('PACKAGE.json', json.dumps({'platform': 'windows', 'architecture': architecture,
+                'status': 'optional unsigned CLI/MCP client; install desktop for model setup', 'files': records}, indent=2))
+        os.replace(temporary_path, output)
+    finally:
+        temporary_path.unlink(missing_ok=True)
+    return architecture
+
+
+def main():
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument('--executable', type=Path, required=True)
+    parser.add_argument('--resources', type=Path, default=ROOT / 'apps/desktop/src-tauri/resources')
+    parser.add_argument('--output', type=Path, default=ROOT / '.searchi/packages/searchi-cli_0.1.0_x64-013.zip')
+    args = parser.parse_args()
+    print(f"Packaged Windows {package(ROOT, args.executable, args.resources, args.output)} CLI: {args.output}")
+
+
+if __name__ == '__main__':
+    main()
+

tools/test_package_windows_cli.py

@@ -0,0 +1,64 @@
+import hashlib
+import json
+from pathlib import Path
+import tempfile
+import unittest
+import zipfile
+
+from package_windows_cli import package
+from test_package_windows import executable
+import stage_extraction_worker as extraction
+
+
+class CliPackageTests(unittest.TestCase):
+    def fixture(self, directory):
+        root = Path(directory)
+        (root / 'Cargo.toml').write_text('workspace')
+        (root / 'Cargo.lock').write_text('lock')
+        for crate in ('searchi-core', 'searchi-media', 'searchi-extract-worker'):
+            base = root / 'crates' / crate
+            (base / 'src').mkdir(parents=True)
+            (base / 'Cargo.toml').write_text(crate)
+            (base / 'src' / ('main.rs' if crate == 'searchi-extract-worker' else 'lib.rs')).write_text('// fixture')
+        (root / 'LICENSE').write_text('license')
+        cli = root / 'cli.exe'
+        helper = root / 'helper.exe'
+        executable(cli)
+        executable(helper)
+        resources = root / 'resources'
+        extraction.stage(helper, extraction.fingerprint(root), resources, root)
+        return root, cli, resources, root / 'client.zip'
+
+    def test_integrity_and_minimal_contents(self):
+        with tempfile.TemporaryDirectory() as directory:
+            root, cli, resources, output = self.fixture(directory)
+            self.assertEqual(package(root, cli, resources, output), 'x86_64')
+            with zipfile.ZipFile(output) as archive:
+                self.assertEqual(set(archive.namelist()), {'searchi-cli.exe', 'searchi-extract-worker.exe',
+                    'extraction-worker.json', 'LICENSE.txt', 'PACKAGE.json'})
+                manifest = json.loads(archive.read('PACKAGE.json'))
+                for record in manifest['files']:
+                    data = archive.read(record['path'])
+                    self.assertEqual(len(data), record['bytes'])
+                    self.assertEqual(hashlib.sha256(data).hexdigest(), record['sha256'])
+
+    def test_rejected_helper_preserves_output(self):
+        for mutation in ('missing', 'tampered', 'architecture'):
+            with self.subTest(mutation=mutation), tempfile.TemporaryDirectory() as directory:
+                root, cli, resources, output = self.fixture(directory)
+                output.write_bytes(b'previous')
+                helper = resources / 'searchi-extract-worker.exe'
+                if mutation == 'missing':
+                    helper.unlink()
+                elif mutation == 'tampered':
+                    helper.write_bytes(helper.read_bytes() + b'changed')
+                else:
+                    executable(cli, 0xAA64)
+                with self.assertRaises(ValueError):
+                    package(root, cli, resources, output)
+                self.assertEqual(output.read_bytes(), b'previous')
+                self.assertEqual(list(root.glob('*.partial')), [])
+
+
+if __name__ == '__main__':
+    unittest.main()