[codex] Skip empty org repos during incremental fetch

src/review_classification/cli/app.py

@@ -51,6 +51,7 @@ def _fetch_repo(
     reset_db: bool,
     verbose: bool,
     incremental: bool = False,
+    skip_empty_incremental: bool = False,
 ) -> None:
     """Fetch and store PRs for a single repository."""
     repo = GitHubRepo.from_string(repo_name)
@@ -75,6 +76,13 @@ def _fetch_repo(
                     f"Incremental fetch: Found latest PR date {start_date} in database."
                 )
         else:
+            if skip_empty_incremental:
+                typer.echo(
+                    f"Skipping {full_name}: no existing PR data found for "
+                    "incremental fetch.",
+                    err=True,
+                )
+                return
             typer.echo(
                 f"Error: Cannot use --incremental for {full_name} "
                 "on an initial fetch or empty database.",
@@ -307,6 +315,7 @@ def add_repo(rc: RepoConfig) -> None:
                     # Create a RepoConfig inheriting from org_cfg, then global
                     rc = RepoConfig(
                         name=r,
+                        from_organization=True,
                         collate_start=(
                             org_cfg.collate_start
                             if org_cfg.collate_start is not None
@@ -353,6 +362,7 @@ def add_repo(rc: RepoConfig) -> None:
         for r in org_repos:
             rc = RepoConfig(
                 name=r,
+                from_organization=True,
                 collate_start=default_collate_start,
                 collate_end=default_collate_end,
                 threshold=default_threshold,
@@ -456,6 +466,7 @@ def fetch(
                 reset_db,
                 verbose,
                 incremental=incremental,
+                skip_empty_incremental=target.from_organization,
             )
 
     except (FileNotFoundError, ValueError) as e:

src/review_classification/cli/config.py

@@ -16,6 +16,7 @@ class RepoConfig:
     min_samples: int | None = None
     start: str | None = None
     end: str | None = None
+    from_organization: bool = False
 
 
 @dataclass
@@ -63,6 +64,7 @@ def resolve(self, repo: RepoConfig) -> RepoConfig:
         """
         return RepoConfig(
             name=repo.name,
+            from_organization=repo.from_organization,
             collate_start=(
                 repo.collate_start
                 if repo.collate_start is not None

tests/cli/test_fetch.py

@@ -0,0 +1,102 @@
+"""Tests for fetch command behavior."""
+
+from datetime import UTC, datetime
+from importlib import import_module
+from pathlib import Path
+from typing import Any
+
+import pytest
+import typer
+
+cli_app = import_module("review_classification.cli.app")
+
+
+def _org_repos(_org_name: str) -> list[str]:
+    return ["my-org/existing", "my-org/new"]
+
+
+def _save_pr(_pr: Any) -> None:
+    return None
+
+
+def _no_latest_pr_date(_repo_name: str) -> None:
+    return None
+
+
+def test_incremental_org_fetch_skips_repo_without_existing_data(
+    monkeypatch: pytest.MonkeyPatch,
+    capsys: pytest.CaptureFixture[str],
+) -> None:
+    """Org incremental fetches skip newly discovered repos without DB history."""
+    fetched_repos: list[str] = []
+
+    def latest_pr_date(repo_name: str) -> datetime | None:
+        if repo_name == "my-org/existing":
+            return datetime(2024, 1, 15, tzinfo=UTC)
+        return None
+
+    def fetch_prs(
+        repo_name: str,
+        start_date: str | None = None,
+        end_date: str | None = None,
+        token: str | None = None,
+    ) -> list[Any]:
+        del start_date, end_date, token
+        fetched_repos.append(repo_name)
+        return []
+
+    monkeypatch.setenv("GITHUB_TOKEN", "test-token")
+    monkeypatch.setattr(
+        "review_classification.queries.github_client.get_org_repos", _org_repos
+    )
+    monkeypatch.setattr(
+        "review_classification.sqlite.database.get_latest_pr_date", latest_pr_date
+    )
+    monkeypatch.setattr(
+        "review_classification.queries.github_client.fetch_prs", fetch_prs
+    )
+    monkeypatch.setattr("review_classification.sqlite.database.init_db", lambda: None)
+    monkeypatch.setattr("review_classification.sqlite.database.save_pr", _save_pr)
+
+    cli_app.fetch(
+        repo=None,
+        org=["my-org"],
+        config=None,
+        collate_start=None,
+        collate_end=None,
+        reset_db=False,
+        verbose=False,
+        incremental=True,
+    )
+
+    captured = capsys.readouterr()
+    assert fetched_repos == ["my-org/existing"]
+    assert (
+        "Skipping my-org/new: no existing PR data found for incremental fetch."
+        in captured.err
+    )
+
+
+def test_incremental_repo_fetch_still_errors_without_existing_data(
+    monkeypatch: pytest.MonkeyPatch,
+    tmp_path: Path,
+) -> None:
+    """Explicit repo incremental fetches still require existing DB history."""
+    monkeypatch.chdir(tmp_path)
+    monkeypatch.setattr(
+        "review_classification.sqlite.database.get_latest_pr_date", _no_latest_pr_date
+    )
+
+    with pytest.raises(typer.Exit) as exc_info:
+        cli_app.fetch(
+            repo=["my-org/new"],
+            org=None,
+            config=None,
+            collate_start=None,
+            collate_end=None,
+            reset_db=False,
+            verbose=False,
+            incremental=True,
+        )
+
+    assert exc_info.value.exit_code == 1