Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.
Python aicode-complexitycode-healthcode-intelligencecode-qualitydead-codedeveloper-toolsdocumentationgit-analyticsmcpopen-sourcepythonrefactoringstatic-analysistechnical-debt
fix(mcp): stop get_change_risk sharing a key name with a different question (#1884)
Raghav Chamadiya35d4848e7f93
diff --git a/docs/agent/MCP_TOOLS.md b/docs/agent/MCP_TOOLS.mdindex 247dd82fb..cf2497a35 100644--- a/docs/agent/MCP_TOOLS.md+++ b/docs/agent/MCP_TOOLS.md@@ -511,13 +511,13 @@ so a narrower set than `get_risk`'s file-level `tests_to_run` — same field nambecause it is the same concern). It is capped at ten, with `total` and`truncated` reporting the overflow and the tail written to the omission store:on `truncated: true` the response carries an `omission_marker`, and-`repowise expand <ref>` returns the rest. Its `missing_tests` buckets flag+`repowise expand <ref>` returns the rest. Its `line_coverage` buckets flag`untested_changes` (covered file, uncovered change), `stale_test_candidates`(covered lines whose guarding test file is absent from the diff), `covered`, and`no_coverage_data` (files absent from the map). When no map is ingested thechange is never reported as untested: `status` becomes `inferred` when theimport graph can name test files reaching the change (candidates, file-level, no-line attribution, and `missing_tests` stays empty because reaching cannot speak+line attribution, and `line_coverage` stays empty because reaching cannot speakto lines), and `no_map` ("run the full suite") when it cannot. `basis` carriesthe same distinction in one word: `measured`, `inferred`, or absent. Build themeasured map with `coverage run --contexts=test` followed bydiff --git a/docs/layers/TEST_INTELLIGENCE.md b/docs/layers/TEST_INTELLIGENCE.mdindex 5de31161c..c0603974e 100644--- a/docs/layers/TEST_INTELLIGENCE.md+++ b/docs/layers/TEST_INTELLIGENCE.md@@ -213,7 +213,7 @@ changed *lines*, so it is a strictly narrower and more useful set:"tests": ["tests/test_auth.py::test_login", "..."],"total": 23,"truncated": true,- "missing_tests": {+ "line_coverage": {"untested_changes": [{"source_file": "...", "uncovered_lines": [...]}],"stale_test_candidates": [...],"covered": [...],@@ -223,7 +223,7 @@ changed *lines*, so it is a strictly narrower and more useful set:}```-The `missing_tests` buckets are the honest breakdown: `untested_changes` is the+The `line_coverage` buckets are the honest breakdown: `untested_changes` is thestrong signal (the file *is* in the map, but nothing covers the lines youtouched), `stale_test_candidates` flags covered lines whose guarding test file isabsent from the diff, and `no_coverage_data` means the file is simply not in thediff --git a/packages/core/src/repowise/core/analysis/change_risk/model.py b/packages/core/src/repowise/core/analysis/change_risk/model.pyindex c720365ab..10622db65 100644--- a/packages/core/src/repowise/core/analysis/change_risk/model.py+++ b/packages/core/src/repowise/core/analysis/change_risk/model.py@@ -138,7 +138,10 @@ SCORE_UNIT = "per-commit"#: blamed by a later fix. Ranking changes by lines-added alone beats the fitted#: model on that benchmark. So the score is reported as what it demonstrably is —#: a diff-size statistic — and ``fix_history`` carries the part that is not.-SCORE_MEASURES = "diff size and spread; not where the change lands"+SCORE_MEASURES = (+ "diff size and spread; not where the change lands - overall_risk_score "+ "is the separate 0-10 for that"+)def _level(score: float) -> str:diff --git a/packages/server/src/repowise/server/mcp_server/tool_change_risk.py b/packages/server/src/repowise/server/mcp_server/tool_change_risk.pyindex 7604479a5..0446de735 100644--- a/packages/server/src/repowise/server/mcp_server/tool_change_risk.py+++ b/packages/server/src/repowise/server/mcp_server/tool_change_risk.py@@ -77,7 +77,7 @@ async def get_change_risk(calibrated per commit; ``risk_percentile`` ranks that shape against recentcommits.- ``impacted_tests.tests_to_run`` names the tests to run; ``missing_tests``+ ``impacted_tests.tests_to_run`` names the tests to run; ``line_coverage``the uncovered lines. ``basis``: ``measured`` = a coverage map proves thosetests run the changed *lines*; ``inferred`` = test files the graph showsreaching it (file-level candidates). No map is never "untested":@@ -366,7 +366,7 @@ def _empty_impacted(status: str, summary: str) -> dict[str, Any]:"tests_to_run": [],"total": 0,"truncated": False,- "missing_tests": {+ "line_coverage": {"untested_changes": [],"stale_test_candidates": [],"covered": [],@@ -610,7 +610,7 @@ async def _inferred_impacted(measured answer.Deliberately file-level and line-blind. Reaching carries no line- attribution, so ``missing_tests`` stays empty rather than being filled from+ attribution, so ``line_coverage`` stays empty rather than being filled froma signal that cannot speak to lines - the distinction this whole blockexists to keep."""@@ -707,7 +707,7 @@ async def _impacted_tests_block("tests_to_run": _cap_tests(tests, collector, "measured"),"total": total,"truncated": total > _IMPACTED_TESTS_LIMIT,- "missing_tests": _serialize_missing(report),+ "line_coverage": _serialize_missing(report),"summary": (f"{total} test(s) cover the changed lines"+ (f"; showing first {_IMPACTED_TESTS_LIMIT}" if total > _IMPACTED_TESTS_LIMIT else "")diff --git a/tests/unit/server/mcp/test_change_risk.py b/tests/unit/server/mcp/test_change_risk.pyindex 49b34dfdc..ce26324da 100644--- a/tests/unit/server/mcp/test_change_risk.py+++ b/tests/unit/server/mcp/test_change_risk.py@@ -221,7 +221,7 @@ async def test_impacted_tests_line_precise_hit_and_miss(tmp_path, monkeypatch) -assert it["total"] == 1assert it["truncated"] is False- mt = it["missing_tests"]+ mt = it["line_coverage"]# other.py is in the map but its changed lines (1,2,3) are uncovered.assert mt["untested_changes"] == [{"source_file": "src/other.py", "uncovered_lines": [1, 2, 3], "changed_line_count": 3}@@ -256,7 +256,7 @@ async def test_impacted_tests_no_map_is_unknown_not_untested(tmp_path, monkeypatassert it["map_present"] is Falseassert it["tests_to_run"] == []# Honest degradation: no untested claim, a "run the suite" summary instead.- assert it["missing_tests"]["untested_changes"] == []+ assert it["line_coverage"]["untested_changes"] == []assert "run the full suite" in it["summary"]@@ -266,7 +266,7 @@ async def test_impacted_tests_falls_back_to_the_graph_without_a_map(tmp_path, moReplaces a bare "run the full suite" with a candidate list on every repothat never ingested a report. Labelled ``inferred`` and file-level: reaching- carries no line attribution, so ``missing_tests`` stays empty rather than+ carries no line attribution, so ``line_coverage`` stays empty rather thanbeing filled from a signal that cannot speak to lines."""from repowise.core.persistence.models import GraphEdge, GraphNode@@ -305,7 +305,7 @@ async def test_impacted_tests_falls_back_to_the_graph_without_a_map(tmp_path, moassert it["basis"] == "inferred"assert it["map_present"] is Falseassert it["tests_to_run"] == ["tests/test_round_trips.py"]- assert it["missing_tests"]["untested_changes"] == []+ assert it["line_coverage"]["untested_changes"] == []# Answered by the import tier: there is no call edge here, which is exactly# when the weaker tier is allowed to speak.assert "reach the changed files in the graph" in it["summary"]@@ -362,3 +362,32 @@ async def test_impacted_tests_no_session_factory_degrades_to_no_index(tmp_path,assert result["impacted_tests"]["status"] == "no_index"assert result["impacted_tests"]["map_present"] is False+++async def test_the_two_risk_tools_do_not_share_a_key_for_different_questions() -> None:+ """``missing_tests`` asked two different questions under one name.++ get_risk's is a list of file paths with no test; this one bucketed changed+ files by line coverage. An agent that learned one misread the other.+ """+ import importlib++ module = importlib.import_module("repowise.server.mcp_server.tool_change_risk")+ empty = module._empty_impacted("no_map", "run the suite")++ assert "missing_tests" not in empty+ assert set(empty["line_coverage"]) == {+ "untested_changes",+ "stale_test_candidates",+ "covered",+ "no_coverage_data",+ }+++def test_score_measures_names_the_other_zero_to_ten() -> None:+ """Whichever risk tool an agent calls first, it learns the other is not it."""+ from repowise.core.analysis.change_risk import SCORE_MEASURES++ assert "diff size and spread" in SCORE_MEASURES+ assert "overall_risk_score" in SCORE_MEASURES