code-mower 0.5.0b27__tar.gz → 0.5.0b29__tar.gz

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: code-mower
-Version: 0.5.0b27
+Version: 0.5.0b29
 Summary: Multi-reviewer AI code audit orchestration
 License-Expression: Apache-2.0
 Requires-Python: >=3.11
@@ -101,11 +101,11 @@ Code Mower currently targets Python 3.11+; Python 3.12 is recommended.
 
 ```bash
 python3.12 --version
-pipx install --python python3.12 code-mower==0.5.0b27
+pipx install --python python3.12 code-mower==0.5.0b29
 code-mower --version
 ```
 
-`0.5.0b27` is a beta release. Until Code Mower publishes a stable `1.0`
+`0.5.0b29` is a beta release. Until Code Mower publishes a stable `1.0`
 line, use the explicit beta version above or allow prereleases with:
 
 ```bash
@@ -148,7 +148,7 @@ First-time command map: [docs/launch-command-surface.md](docs/launch-command-sur
 
 The current PyPI beta has been rehearsed end-to-end from a clean install:
 [First-User Install Rehearsal](docs/first-user-install-rehearsal.md) records
-the latest 10/10 public-package readiness proof for `code-mower==0.5.0b27`.
+the latest 10/10 public-package readiness proof for `code-mower==0.5.0b29`.
 
 ## Why Not Just Run Codex Or Claude Yourself?
 
@@ -253,9 +253,9 @@ best?" to "which AI builder plus reviewer loop ships best on this product?" See
 
 ## Installation Status
 
-The current public beta is `v0.5.0-beta.27` from
+The current public beta is `v0.5.0-beta.29` from
 [codemower-ai/code-mower](https://github.com/codemower-ai/code-mower), published
-as `code-mower==0.5.0b27` on [PyPI](https://pypi.org/project/code-mower/).
+as `code-mower==0.5.0b29` on [PyPI](https://pypi.org/project/code-mower/).
 GitHub releases remain the auditable source for tags, build artifacts, and
 release notes.
 

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/README.md

@@ -86,11 +86,11 @@ Code Mower currently targets Python 3.11+; Python 3.12 is recommended.
 
 ```bash
 python3.12 --version
-pipx install --python python3.12 code-mower==0.5.0b27
+pipx install --python python3.12 code-mower==0.5.0b29
 code-mower --version
 ```
 
-`0.5.0b27` is a beta release. Until Code Mower publishes a stable `1.0`
+`0.5.0b29` is a beta release. Until Code Mower publishes a stable `1.0`
 line, use the explicit beta version above or allow prereleases with:
 
 ```bash
@@ -133,7 +133,7 @@ First-time command map: [docs/launch-command-surface.md](docs/launch-command-sur
 
 The current PyPI beta has been rehearsed end-to-end from a clean install:
 [First-User Install Rehearsal](docs/first-user-install-rehearsal.md) records
-the latest 10/10 public-package readiness proof for `code-mower==0.5.0b27`.
+the latest 10/10 public-package readiness proof for `code-mower==0.5.0b29`.
 
 ## Why Not Just Run Codex Or Claude Yourself?
 
@@ -238,9 +238,9 @@ best?" to "which AI builder plus reviewer loop ships best on this product?" See
 
 ## Installation Status
 
-The current public beta is `v0.5.0-beta.27` from
+The current public beta is `v0.5.0-beta.29` from
 [codemower-ai/code-mower](https://github.com/codemower-ai/code-mower), published
-as `code-mower==0.5.0b27` on [PyPI](https://pypi.org/project/code-mower/).
+as `code-mower==0.5.0b29` on [PyPI](https://pypi.org/project/code-mower/).
 GitHub releases remain the auditable source for tags, build artifacts, and
 release notes.
 

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "code-mower"
-version = "0.5.0b27"
+version = "0.5.0b29"
 description = "Multi-reviewer AI code audit orchestration"
 requires-python = ">=3.11"
 readme = "README.md"

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/src/code_mower/__init__.py

@@ -1,3 +1,3 @@
 """Code Mower package."""
 
-__version__ = "0.5.0b27"
+__version__ = "0.5.0b29"

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/src/code_mower/cli.py

@@ -14,7 +14,7 @@ from pathlib import Path
 if __package__ in {None, ""}:
     raise SystemExit(
         "code_mower.cli is a packaged entrypoint. Install Code Mower with "
-        "`pipx install code-mower==0.5.0b27`, or run source checkouts with "
+        "`pipx install code-mower==0.5.0b29`, or run source checkouts with "
         "`PYTHONPATH=src python -m code_mower.cli`."
     )
 
@@ -175,6 +175,7 @@ def _providers_main(argv: list[str]) -> int:
             print(f"Status: {report['status']}")
             print(f"Providers: {report['provider_count']}")
             print(f"Missing model env: {report['missing_model_env_count']}")
+            print(f"Missing CLI version probes: {report['missing_version_probe_count']}")
             for row in report["providers"]:
                 if row["action"] == "set_model_env":
                     aliases = ", ".join(row["env_names"]) or "none"
@@ -182,6 +183,8 @@ def _providers_main(argv: list[str]) -> int:
                         f"- {row['lane_id']}: set {row['preferred_env']} "
                         f"(aliases: {aliases})"
                     )
+                elif row["driver"] == "local_cli" and row["version_source"] == "missing":
+                    print(f"- {row['lane_id']}: CLI version probe is missing")
                 elif row["model_source"] == "vendor_hidden":
                     print(f"- {row['lane_id']}: provider hides model/version metadata")
                 elif row["model_source"] == "default":

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/src/code_mower/cloud_client/doctor.py

@@ -183,19 +183,40 @@ def run_cloud_doctor(
             )
             if isinstance(provenance, dict):
                 missing_model_events = int(
+                    provenance.get(
+                        "benchmark_events_missing_model_provenance",
+                        provenance.get("events_missing_model_provenance"),
+                    )
+                    or 0
+                )
+                raw_missing_model_events = int(
                     provenance.get("events_missing_model_provenance") or 0
                 )
                 provenance_tools = provenance.get("tools", [])
                 if not isinstance(provenance_tools, list):
                     provenance_tools = []
-                missing_model_tools = sorted(
-                    {
-                        str(tool.get("provider") or tool.get("tool_name") or "unknown")
-                        for tool in provenance_tools
-                        if isinstance(tool, dict)
-                        and "missing" in set(tool.get("model_sources") or [])
-                    }
+                benchmark_missing_providers = provenance.get(
+                    "benchmark_model_missing_providers"
                 )
+                if isinstance(benchmark_missing_providers, list):
+                    missing_model_tools = sorted(
+                        {
+                            str(provider).strip()
+                            for provider in benchmark_missing_providers
+                            if str(provider).strip()
+                        }
+                    )
+                else:
+                    missing_model_tools = sorted(
+                        {
+                            str(
+                                tool.get("provider") or tool.get("tool_name") or "unknown"
+                            )
+                            for tool in provenance_tools
+                            if isinstance(tool, dict)
+                            and "missing" in set(tool.get("model_sources") or [])
+                        }
+                    )
                 model_env_by_provider = _provider_model_env_names(missing_model_tools)
                 model_env_examples = sorted(
                     {
@@ -236,10 +257,13 @@ def run_cloud_doctor(
                             "name": "model-provenance",
                             "status": "warn",
                             "message": (
-                                f"{missing_model_events} events are missing model provenance"
+                                f"{missing_model_events} benchmark evidence events "
+                                "are missing model provenance"
                             ),
                             "detail": {
                                 "providers": missing_model_tools,
+                                "raw_missing_model_events": raw_missing_model_events,
+                                "benchmark_missing_model_events": missing_model_events,
                                 "model_env_by_provider": model_env_by_provider,
                                 "model_env_examples": model_env_examples,
                                 "model_env_commands": model_env_commands,
@@ -253,7 +277,14 @@ def run_cloud_doctor(
                         {
                             "name": "model-provenance",
                             "status": "pass",
-                            "message": "bundle has model provenance for all structured events",
+                            "message": (
+                                "bundle has model provenance for all benchmark evidence "
+                                "events"
+                            ),
+                            "detail": {
+                                "raw_missing_model_events": raw_missing_model_events,
+                                "benchmark_missing_model_events": missing_model_events,
+                            },
                         }
                     )
         except CloudBundleError as exc:

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/src/code_mower/cloud_client/export.py

@@ -20,6 +20,15 @@ from .errors import CloudBundleError
 from .events import normalize_event, safe_kind
 
 
+INVENTORY_ONLY_PROVENANCE_EVENT_TYPES = {"provider_catalog_snapshot"}
+
+
+def is_benchmark_provenance_event(event: dict[str, Any]) -> bool:
+    """Return whether an event should count toward benchmark provenance coverage."""
+
+    return str(event.get("event_type") or "") not in INVENTORY_ONLY_PROVENANCE_EVENT_TYPES
+
+
 def safe_target_name(path: Path, index: int, *, anonymous: bool = False) -> str:
     suffix = path.suffix.lower()
     if suffix not in {".json", ".jsonl", ".md", ".txt"}:
@@ -245,12 +254,25 @@ def build_provenance_summary(events: list[dict[str, Any]]) -> dict[str, Any]:
     missing_tool_events = 0
     missing_model_events = 0
     missing_tool_version_events = 0
+    benchmark_event_count = 0
+    benchmark_missing_model_events = 0
+    benchmark_missing_tool_version_events = 0
+    benchmark_model_missing_providers: set[str] = set()
     for event in events:
+        is_benchmark_event = is_benchmark_provenance_event(event)
+        if is_benchmark_event:
+            benchmark_event_count += 1
         tool = event.get("tool")
         if not isinstance(tool, dict):
             missing_tool_events += 1
             missing_model_events += 1
             missing_tool_version_events += 1
+            if is_benchmark_event:
+                benchmark_missing_model_events += 1
+                benchmark_missing_tool_version_events += 1
+                benchmark_model_missing_providers.add(
+                    str(event.get("provider") or "unknown").strip() or "unknown"
+                )
             continue
         tool_name = str(tool.get("tool_name") or "").strip()
         provider = str(tool.get("provider") or event.get("provider") or "").strip()
@@ -262,13 +284,24 @@ def build_provenance_summary(events: list[dict[str, Any]]) -> dict[str, Any]:
             missing_tool_events += 1
             missing_model_events += 1
             missing_tool_version_events += 1
+            if is_benchmark_event:
+                benchmark_missing_model_events += 1
+                benchmark_missing_tool_version_events += 1
+                benchmark_model_missing_providers.add("unknown")
             continue
         if not model and not model_version_raw and model_source in {"", "missing"}:
             missing_model_events += 1
+            if is_benchmark_event:
+                benchmark_missing_model_events += 1
+                benchmark_model_missing_providers.add(
+                    provider or tool_name or "unknown"
+                )
         if not tool_version:
             version_source = str(tool.get("version_source") or "").strip()
             if version_source in {"", "missing"}:
                 missing_tool_version_events += 1
+                if is_benchmark_event:
+                    benchmark_missing_tool_version_events += 1
         key = (tool_name or "unknown", provider or "unknown", model or "")
         row = tools.setdefault(
             key,
@@ -328,6 +361,15 @@ def build_provenance_summary(events: list[dict[str, Any]]) -> dict[str, Any]:
         "events_with_tool_version_provenance": len(events)
         - missing_tool_version_events,
         "events_missing_tool_version_provenance": missing_tool_version_events,
+        "inventory_event_count": len(events) - benchmark_event_count,
+        "benchmark_event_count": benchmark_event_count,
+        "benchmark_events_with_model_provenance": benchmark_event_count
+        - benchmark_missing_model_events,
+        "benchmark_events_missing_model_provenance": benchmark_missing_model_events,
+        "benchmark_events_with_tool_version_provenance": benchmark_event_count
+        - benchmark_missing_tool_version_events,
+        "benchmark_events_missing_tool_version_provenance": benchmark_missing_tool_version_events,
+        "benchmark_model_missing_providers": sorted(benchmark_model_missing_providers),
         "tools": normalized_tools,
     }
 

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/src/code_mower/providers/provenance.py

@@ -235,7 +235,12 @@ def build_provider_model_env_report(
             matched.add(provider)
         config = _provider_config(lane)
         driver = _safe_text(_lane_value(lane, "driver")) or "unknown"
-        model, model_source = _configured_model(config, driver)
+        tool, provenance_detail = build_provider_lane_tool_provenance(
+            lane_id,
+            lane,
+            source="code-mower providers provenance-env",
+        )
+        model_source = _safe_text(provenance_detail.get("model_source"))
         env_names = model_env_names(config)
         preferred_env = preferred_model_env_name(config)
         env_status = [
@@ -256,8 +261,16 @@ def build_provider_model_env_report(
                 "lane_id": lane_id,
                 "provider": provider,
                 "driver": driver,
-                "model_known": bool(model),
+                "model_known": bool(provenance_detail.get("model_known")),
                 "model_source": model_source,
+                "tool_name": tool.get("tool_name", ""),
+                "tool_version": tool.get("tool_version", ""),
+                "version_known": bool(provenance_detail.get("version_known")),
+                "version_source": _safe_text(
+                    provenance_detail.get("version_source")
+                ),
+                "command": _safe_text(provenance_detail.get("command")),
+                "command_found": bool(provenance_detail.get("command_found")),
                 "env_names": list(env_names),
                 "preferred_env": preferred_env,
                 "env_status": env_status,
@@ -267,11 +280,17 @@ def build_provider_model_env_report(
         )
     unknown = sorted(requested - matched)
     missing = [row for row in rows if row["action"] == "set_model_env"]
+    missing_version_probe = [
+        row
+        for row in rows
+        if row["driver"] == "local_cli" and row["version_source"] == "missing"
+    ]
     return {
         "mode": "code-mower-provider-model-env-report",
         "providers": rows,
         "provider_count": len(rows),
         "missing_model_env_count": len(missing),
+        "missing_version_probe_count": len(missing_version_probe),
         "unknown_providers": unknown,
         "status": "fail" if unknown else "pass",
     }

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/src/code_mower.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: code-mower
-Version: 0.5.0b27
+Version: 0.5.0b29
 Summary: Multi-reviewer AI code audit orchestration
 License-Expression: Apache-2.0
 Requires-Python: >=3.11
@@ -101,11 +101,11 @@ Code Mower currently targets Python 3.11+; Python 3.12 is recommended.
 
 ```bash
 python3.12 --version
-pipx install --python python3.12 code-mower==0.5.0b27
+pipx install --python python3.12 code-mower==0.5.0b29
 code-mower --version
 ```
 
-`0.5.0b27` is a beta release. Until Code Mower publishes a stable `1.0`
+`0.5.0b29` is a beta release. Until Code Mower publishes a stable `1.0`
 line, use the explicit beta version above or allow prereleases with:
 
 ```bash
@@ -148,7 +148,7 @@ First-time command map: [docs/launch-command-surface.md](docs/launch-command-sur
 
 The current PyPI beta has been rehearsed end-to-end from a clean install:
 [First-User Install Rehearsal](docs/first-user-install-rehearsal.md) records
-the latest 10/10 public-package readiness proof for `code-mower==0.5.0b27`.
+the latest 10/10 public-package readiness proof for `code-mower==0.5.0b29`.
 
 ## Why Not Just Run Codex Or Claude Yourself?
 
@@ -253,9 +253,9 @@ best?" to "which AI builder plus reviewer loop ships best on this product?" See
 
 ## Installation Status
 
-The current public beta is `v0.5.0-beta.27` from
+The current public beta is `v0.5.0-beta.29` from
 [codemower-ai/code-mower](https://github.com/codemower-ai/code-mower), published
-as `code-mower==0.5.0b27` on [PyPI](https://pypi.org/project/code-mower/).
+as `code-mower==0.5.0b29` on [PyPI](https://pypi.org/project/code-mower/).
 GitHub releases remain the auditable source for tags, build artifacts, and
 release notes.
 

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/tests/test_cloud_client.py

@@ -279,6 +279,11 @@ def test_provenance_summary_treats_vendor_hidden_model_as_known_source() -> None
     assert summary["events_missing_model_provenance"] == 0
     assert summary["events_with_tool_version_provenance"] == 1
     assert summary["events_missing_tool_version_provenance"] == 0
+    assert summary["inventory_event_count"] == 1
+    assert summary["benchmark_event_count"] == 0
+    assert summary["benchmark_events_with_model_provenance"] == 0
+    assert summary["benchmark_events_missing_model_provenance"] == 0
+    assert summary["benchmark_model_missing_providers"] == []
     assert summary["tools"][0]["model_sources"] == ["vendor_hidden"]
     assert summary["tools"][0]["version_sources"] == ["not_probed"]
 
@@ -299,7 +304,7 @@ def test_provenance_summary_preserves_source_quality_fields() -> None:
                     "tool": {
                         "role": "reporter",
                         "tool_name": "code-mower",
-                        "tool_version": "0.5.0b27",
+                        "tool_version": "0.5.0b29",
                         "provider": "code-mower",
                         "model": "",
                         "model_source": "not_applicable",
@@ -337,6 +342,13 @@ def test_provenance_summary_preserves_source_quality_fields() -> None:
     assert summary["events_missing_model_provenance"] == 1
     assert summary["events_with_tool_version_provenance"] == 2
     assert summary["events_missing_tool_version_provenance"] == 0
+    assert summary["inventory_event_count"] == 1
+    assert summary["benchmark_event_count"] == 1
+    assert summary["benchmark_events_with_model_provenance"] == 1
+    assert summary["benchmark_events_missing_model_provenance"] == 0
+    assert summary["benchmark_events_with_tool_version_provenance"] == 1
+    assert summary["benchmark_events_missing_tool_version_provenance"] == 0
+    assert summary["benchmark_model_missing_providers"] == []
     rows = {row["tool_name"]: row for row in summary["tools"]}
     assert rows["code-mower"]["model_sources"] == ["not_applicable"]
     assert rows["code-mower"]["version_sources"] == ["package_version"]
@@ -460,6 +472,49 @@ def test_cloud_doctor_warns_when_model_provenance_is_missing() -> None:
         assert "CODE_MOWER_CODEX_MODEL" in check["remediation"]
 
 
+def test_cloud_doctor_does_not_warn_for_inventory_only_model_gaps() -> None:
+    with tempfile.TemporaryDirectory() as tmp:
+        root = Path(tmp)
+        build_cloud_bundle(
+            reports=[],
+            events=[
+                {
+                    "event_type": "provider_catalog_snapshot",
+                    "repo_slug": "owner/repo",
+                    "provider": "gemini",
+                    "lens": "base",
+                    "status": "observed",
+                    "dimensions": {"catalog_snapshot": True},
+                    "tool": {
+                        "role": "reviewer",
+                        "tool_name": "gemini",
+                        "provider": "gemini",
+                        "tool_version": "0.45.2",
+                        "model_source": "missing",
+                    },
+                }
+            ],
+            output_dir=root / "bundle",
+            repo_slug="owner/repo",
+        )
+
+        report = run_cloud_doctor(
+            bundle_dir=root / "bundle",
+            endpoint="https://codemower.com/api/ingest",
+            token_env="CODE_MOWER_TEST_CLOUD_TOKEN",
+            require_token=False,
+        )
+
+        check = next(
+            item for item in report["checks"] if item["name"] == "model-provenance"
+        )
+        assert report["status"] == "pass"
+        assert check["status"] == "pass"
+        assert "benchmark evidence events" in check["message"]
+        assert check["detail"]["raw_missing_model_events"] == 1
+        assert check["detail"]["benchmark_missing_model_events"] == 0
+
+
 def test_cloud_doctor_fans_out_multiple_model_provenance_commands() -> None:
     with tempfile.TemporaryDirectory() as tmp:
         root = Path(tmp)

{code_mower-0.5.0b27 → code_mower-0.5.0b29}/tests/test_provider_local_cli.py

@@ -178,6 +178,65 @@ def test_provider_model_env_report_never_exposes_env_value(monkeypatch) -> None:
     assert row["action"] == "none"
 
 
+def test_provider_model_env_report_includes_cli_version_status(
+    monkeypatch,
+    tmp_path: Path,
+) -> None:
+    executable = _write_executable(
+        tmp_path / "custom-reviewer",
+        "#!/bin/sh\nprintf 'custom-reviewer 2.0.0\\n'\n",
+    )
+    monkeypatch.setenv("PATH", os.fspath(executable.parent))
+    monkeypatch.setenv("CUSTOM_REVIEWER_MODEL", "do-not-render-this-model")
+    lanes = {
+        "custom_reviewer": {
+            "driver": "local_cli",
+            "provider": "custom",
+            "provider_config": {
+                "command": "custom-reviewer",
+                "model_env": "CUSTOM_REVIEWER_MODEL",
+            },
+        }
+    }
+
+    report = build_provider_model_env_report(lanes)
+
+    assert report["missing_model_env_count"] == 0
+    assert report["missing_version_probe_count"] == 0
+    row = report["providers"][0]
+    assert row["tool_name"] == "custom-reviewer"
+    assert row["tool_version"] == "custom-reviewer 2.0.0"
+    assert row["version_known"] is True
+    assert row["version_source"] == "cli_version_probe"
+    assert row["command"] == "custom-reviewer"
+    assert row["command_found"] is True
+    assert "do-not-render-this-model" not in str(report)
+
+
+def test_provider_model_env_report_counts_missing_cli_version_probe(monkeypatch) -> None:
+    monkeypatch.setenv("PATH", "")
+    lanes = {
+        "missing_reviewer": {
+            "driver": "local_cli",
+            "provider": "missing",
+            "provider_config": {
+                "command": "missing-reviewer",
+                "model_env": "MISSING_REVIEWER_MODEL",
+            },
+        }
+    }
+
+    report = build_provider_model_env_report(lanes)
+
+    assert report["missing_model_env_count"] == 1
+    assert report["missing_version_probe_count"] == 1
+    row = report["providers"][0]
+    assert row["version_known"] is False
+    assert row["version_source"] == "missing"
+    assert row["command"] == "missing-reviewer"
+    assert row["command_found"] is False
+
+
 def test_providers_provenance_env_cli_shell_output(monkeypatch, capsys) -> None:
     monkeypatch.delenv("CODE_MOWER_ANTIGRAVITY_MODEL", raising=False)
     monkeypatch.delenv("ANTIGRAVITY_MODEL", raising=False)
@@ -377,11 +436,11 @@ def test_provider_lane_tool_provenance_reads_model_from_selected_profile(
 def test_code_mower_reporter_provenance_marks_model_not_applicable() -> None:
     tool = build_code_mower_tool_provenance(
         source="unit-test",
-        version="0.5.0b27",
+        version="0.5.0b29",
     )
 
     assert tool["tool_name"] == "code-mower"
-    assert tool["tool_version"] == "0.5.0b27"
+    assert tool["tool_version"] == "0.5.0b29"
     assert tool["provider"] == "code-mower"
     assert tool["model"] == ""
     assert tool["model_source"] == "not_applicable"