Skip to content

Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.

fix(health): count files and findings the same way on every surface (#3532)

Raghav Chamadiyaa2e08a6bbf60
diff --git a/packages/cli/src/repowise/cli/commands/health_cmd/command.py b/packages/cli/src/repowise/cli/commands/health_cmd/command.py
index 94c032338..1e3ff713d 100644
--- a/packages/cli/src/repowise/cli/commands/health_cmd/command.py
+++ b/packages/cli/src/repowise/cli/commands/health_cmd/command.py
@@ -37,6 +37,7 @@ from repowise.core.analysis.health.models import split_by_origin
from repowise.core.analysis.health.rows import detail_map, split_unscored
from repowise.core.analysis.health.scope import DEFAULT_SCOPE, SCOPES, parse_scope
from repowise.core.analysis.health.scoring import compute_kpis
+from repowise.core.persistence.crud.analysis.health import open_finding_count
from .codegen import _generate_refactoring_code
from .persist import (
@@ -469,7 +470,8 @@ def health_command(
console.print(table)
if findings:
- console.print(f"\n[bold]{len(findings)}[/bold] marker findings:")
+ # The total get_overview prints; advisory rows are listed, not counted.
+ console.print(f"\n[bold]{open_finding_count(findings)}[/bold] open findings:")
f_table = Table()
f_table.add_column("Severity", style="magenta")
f_table.add_column("Marker", style="cyan")
diff --git a/packages/core/src/repowise/core/persistence/crud/analysis/health.py b/packages/core/src/repowise/core/persistence/crud/analysis/health.py
index e689a3eee..05503e269 100644
--- a/packages/core/src/repowise/core/persistence/crud/analysis/health.py
+++ b/packages/core/src/repowise/core/persistence/crud/analysis/health.py
@@ -4,7 +4,7 @@
from __future__ import annotations
import json
-from collections.abc import Sequence
+from collections.abc import Iterable, Sequence
from datetime import datetime
from pathlib import Path
from typing import TYPE_CHECKING, Any, NamedTuple
@@ -415,7 +415,26 @@ async def backfill_module_attribution(
return changed
-async def _health_exclude_spec(session: AsyncSession, repository_id: str) -> Any:
+def open_finding_count(findings: Iterable[Any]) -> int:
+ """The open findings total every health headline prints.
+
+ Advisory findings carry a zero health impact, so counting them would grow
+ the headline without anything having got worse.
+ """
+ return sum(1 for f in findings if getattr(f, "dimension", None) != ADVISORY_DIMENSION)
+
+
+async def health_exclude_spec(
+ session: AsyncSession, repository_id: str, repo_path: Path | str | None = None
+) -> Any:
+ """The exclude patterns every health surface narrows its files by.
+
+ The repository's settings and ``exclude_patterns`` only. Not the gitignore
+ stack: ingestion keeps a file git tracks even when an ignore rule matches
+ it, so a stored row under such a rule is a real file of the repository.
+ ``repo_path`` is where to read the config when the caller serves a
+ checkout other than the stored ``local_path``.
+ """
repo = await session.get(Repository, repository_id)
if repo is None:
return None
@@ -441,7 +460,7 @@ async def _health_exclude_spec(session: AsyncSession, repository_id: str) -> Any
try:
from repowise.core.repo_config import load_repo_config
- cfg = load_repo_config(Path(repo.local_path))
+ cfg = load_repo_config(Path(repo_path or repo.local_path))
if isinstance(cfg, dict):
_add(cfg.get("exclude_patterns"))
except Exception:
@@ -544,7 +563,7 @@ async def get_health_findings(
result = await session.execute(q)
kept = _filter_excluded_paths(
list(result.scalars().all()),
- await _health_exclude_spec(session, repository_id),
+ await health_exclude_spec(session, repository_id),
)
# A filter that drops some of a function's findings would misread its
# shape, so the tier then reads every open finding on the kept files.
@@ -638,7 +657,7 @@ async def get_health_metrics(
result = await session.execute(q)
rows = _filter_excluded_paths(
list(result.scalars().all()),
- await _health_exclude_spec(session, repository_id),
+ await health_exclude_spec(session, repository_id),
)
if len(rows) < 2:
# Nothing to order. Skips the aggregate for the single-file lookups
@@ -684,7 +703,7 @@ async def get_average_health(session: AsyncSession, repository_id: str) -> float
)
).all()
),
- await _health_exclude_spec(session, repository_id),
+ await health_exclude_spec(session, repository_id),
)
# Files with no score (no health dialect for their language) are skipped,
# so a repository of nothing else reads unmeasured, never 10.0.
@@ -736,7 +755,7 @@ async def _open_finding_totals_rows(session: AsyncSession, repository_id: str) -
.order_by(HealthFinding.health_impact.desc(), HealthFinding.id)
)
return _filter_excluded_paths(
- plain_rows(rows), await _health_exclude_spec(session, repository_id)
+ plain_rows(rows), await health_exclude_spec(session, repository_id)
)
@@ -852,9 +871,7 @@ async def get_health_summary(
"worst_performer_score": round(worst.score, 2),
"worst_test_path": worst_test.file_path if worst_test is not None else None,
"worst_test_score": round(worst_test.score, 2) if worst_test is not None else None,
- # Advisory findings carry a zero health impact, so counting them
- # here would grow the headline without anything having got worse.
- "open_findings": len(findings) - by_dim.get(ADVISORY_DIMENSION, 0),
+ "open_findings": open_finding_count(findings),
"maintainability_average": _rounded(maintainability_average),
"performance_average": _rounded(performance_average),
# The headline's two halves, in deduction points. Both come off columns
diff --git a/packages/server/src/repowise/server/mcp_server/tool_health/population.py b/packages/server/src/repowise/server/mcp_server/tool_health/population.py
index 02293fe1d..cae1c5da7 100644
--- a/packages/server/src/repowise/server/mcp_server/tool_health/population.py
+++ b/packages/server/src/repowise/server/mcp_server/tool_health/population.py
@@ -12,9 +12,10 @@ from repowise.core.analysis.health.counts import project as project_counts
from repowise.core.analysis.health.models import split_by_origin
from repowise.core.analysis.health.rows import scored_rows
from repowise.core.analysis.health.scope import parse_scope
+from repowise.core.persistence.crud.analysis.health import health_exclude_spec
from repowise.core.persistence.crud.analysis.plain_rows import plain_rows
from repowise.core.persistence.models import HealthFileMetric
-from repowise.server.mcp_server._helpers import _get_exclude_spec, filter_rows_by_attr
+from repowise.server.mcp_server._helpers import filter_rows_by_attr
from repowise.server.mcp_server.tool_health.request import HealthRequest
from repowise.server.mcp_server.tool_health.targets import _expand_module_targets
@@ -73,7 +74,10 @@ async def load_population(
all_metrics_q = select(*HealthFileMetric.__table__.columns).where(
HealthFileMetric.repository_id == repository.id
)
- exclude_spec = _get_exclude_spec(repo_path)
+ # The rule the CLI, get_overview and status count files by. The MCP-wide
+ # spec adds the gitignore stack, which hid a tracked file its pattern
+ # matched and left get_health one file short of every other surface.
+ exclude_spec = await health_exclude_spec(session, repository.id, repo_path)
indexed_rows = plain_rows(await session.execute(all_metrics_q))
all_metrics = filter_rows_by_attr(indexed_rows, "file_path", exclude_spec)
# Lets an unresolved target report "excluded" rather than "no_such_path".
diff --git a/tests/unit/server/mcp/test_mcp_exclude_filter.py b/tests/unit/server/mcp/test_mcp_exclude_filter.py
index 059d05771..9844a215b 100644
--- a/tests/unit/server/mcp/test_mcp_exclude_filter.py
+++ b/tests/unit/server/mcp/test_mcp_exclude_filter.py
@@ -147,6 +147,31 @@ async def test_get_health_excludes_configured_paths(setup_mcp, health_data, tmp_
assert "src/auth/service.py" in paths
+async def test_get_health_counts_a_file_an_ignore_rule_matches(
+ setup_mcp, health_data, tmp_path, monkeypatch
+):
+ """Ingestion keeps a tracked file an ignore rule matches, and so do the CLI,
+ get_overview and status. get_health dropped it and reported one file fewer."""
+ import repowise.server.mcp_server as mcp_mod
+ from repowise.server.mcp_server import get_health
+
+ (tmp_path / ".gitignore").write_text("src/db/\n", encoding="utf-8")
+ monkeypatch.setattr(mcp_mod._state, "_repo_path", str(tmp_path))
+
+ result = await get_health(only=["worst_files"])
+ assert "src/db/models.py" in {m["file_path"] for m in result.get("worst_files", [])}
+
+
+def test_open_finding_count_leaves_advisory_out() -> None:
+ from types import SimpleNamespace
+
+ from repowise.core.persistence.crud.analysis.health import open_finding_count
+
+ rows = [SimpleNamespace(dimension=d) for d in ("defect", "advisory", None, "performance")]
+ assert open_finding_count(rows) == 3
+
+
# ---------------------------------------------------------------------------
# Targeted traversal-leak tests (edges / paths / co-change), not just helpers
# ---------------------------------------------------------------------------