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(mcp): treat tree-sitter query and history-tier files as what they are (#3723)

Swati Ahujae167bc570312
diff --git a/docs/reference/MCP_RESPONSE_FIELDS.md b/docs/reference/MCP_RESPONSE_FIELDS.md
index cf72f1cd8..65df2eac6 100644
--- a/docs/reference/MCP_RESPONSE_FIELDS.md
+++ b/docs/reference/MCP_RESPONSE_FIELDS.md
@@ -215,6 +215,7 @@ Per file:
| owners, reviewers | Primary owner (`primary_owner`) and the owner clause in `risk_summary` are unconditional; detailed owner metrics (`owner_pct`, `owner_line_pct`, `recent_owner`, `recent_owner_pct`, `bus_factor`, `contributor_count`) are opt-in under `include=["owners"]`; recommended reviewers stay in PR mode |
| test gaps, `security_signals` | Test coverage gaps and security findings for the file |
| `resolved`, `unresolved_reason` | A target naming no indexed file: `unsupported_target_kind` (a `module:` id), `directory`, `not_indexed`, `no_such_path`. Counts are omitted, never zeroed |
+| `signal_tier`, `signal_tier_reason` | `"history"` on a file indexed for commit history only (docs, config, query and other non-code files): the card carries commit counts, `last_commit_at`, `trend`, owner and `security_signals`, and omits hotspot, co-change, dependents and classification rather than zeroing them |
Opt-in blocks: `owners` adds `owner_pct`, `owner_line_pct`, `recent_owner`, `recent_owner_pct`, `bus_factor` and `contributor_count`. `tests` adds the PR directive's typed `test_recommendations`, and `blast` adds `pr_blast_radius`, the PR dossier behind the directive (transitive files, co-change warnings, test gaps, the structural score; reviewers stay in the directive). `graph` adds typed `dependents` (direct versus transitive), `consumers` (typed contract consumers only), `cross_repo_links`, `impact_surface` and `direct_risks`, with `relationship_analysis` distinguishing an empty analysis from an unavailable or partial one. `churn` adds `change_magnitude`, `risk_type` and `change_pattern`. `scales` adds the unit, range and calibration of every scalar; it is identical on every call, so ask once. A multi-target call also carries `global_hotspots`.
@@ -259,7 +260,7 @@ Every float is rounded to 4 significant digits. Direct rows' `structural_score`
`include=["diagnostics"]` adds `score` (the raw 0 to 10 number), `fallback_band`, `risk_authority`, `score_measures`, `score_unit`, `baseline_sample_size`, `features` and `drivers`, plus `health_delta.analyzer` (analyzer and rules fingerprint), `health_delta.base` and `health_delta.head` (the two resolved revisions) and `health_delta.limits`. `include=["scales"]` adds each field's unit, range and calibration.
-**health_delta.** Both revisions are analysed from their own content, and a finding at head is reported only when the diff explains it. `scope` counts changed, eligible, analysed, skipped and failed files; `status` is `available`, `partial` or `unavailable`, and `partial` is never a clean bill. Documentation and configuration files (Markdown, JSON, YAML, TOML and the other doc or config extensions) are skipped with reason `not_code` and never make a run `partial`; a change made only of them reads `clear_in_analyzed_scope` with a headline saying health analysis does not cover them. HTML, schema and other files with no health dialect still make a run `partial`. `introduced`, `worsened` and `resolved` are totals; `top_findings` holds the three most actionable, with `findings_total`. Each finding names `dimension`, `biomarker`, `severity`, `path`, `symbol`, head-side `lines`, a `reason`, and an `attribution` (`basis` of `added_lines`, `changed_symbol`, `changed_call_edge`, `new_file`, `file_change`, `context_change` or `unknown`, with a `confidence`). Identity ignores line numbers, so moved code introduces nothing. `inspect` gives the `finding_id` call that expands one; ids are bound to the two revisions. A finding that matches a stored one carries `health_reference` for `get_health(finding_id=...)`. Performance findings carry `opportunity_id` and are ordered by opportunity rank.
+**health_delta.** Both revisions are analysed from their own content, and a finding at head is reported only when the diff explains it. `scope` counts changed, eligible, analysed, skipped and failed files; `status` is `available`, `partial` or `unavailable`, and `partial` is never a clean bill. Documentation and configuration files (Markdown, JSON, YAML, TOML and the other doc or config extensions) are skipped with reason `not_code` and never make a run `partial`; a change made only of them reads `clear_in_analyzed_scope` with a headline saying health analysis does not cover them. Tree-sitter query files (`.scm` under a `queries` directory) are skipped with reason `not_analyzable_data`, which also never makes a run `partial`; a change made only of them reads `clear_in_analyzed_scope` with its own headline, since they can still change behaviour. HTML, schema and other files with no health dialect still make a run `partial`. `introduced`, `worsened` and `resolved` are totals; `top_findings` holds the three most actionable, with `findings_total`. Each finding names `dimension`, `biomarker`, `severity`, `path`, `symbol`, head-side `lines`, a `reason`, and an `attribution` (`basis` of `added_lines`, `changed_symbol`, `changed_call_edge`, `new_file`, `file_change`, `context_change` or `unknown`, with a `confidence`). Identity ignores line numbers, so moved code introduces nothing. `inspect` gives the `finding_id` call that expands one; ids are bound to the two revisions. A finding that matches a stored one carries `health_reference` for `get_health(finding_id=...)`. Performance findings carry `opportunity_id` and are ordered by opportunity rank.
**impacted_tests.** The same selection `repowise impacted-tests` makes over the whole change (every touched path, whatever the score's filters). `status` is `selected` or `run_all` (or `no_index`, `unknown`, `config_invalid`, `timeout`, `busy`, `no_source_line_changes` when nothing was selected, each with `run_all` true; `timeout` means the selection took over 30 seconds and `busy` that one for this repository is still running, and `repowise impacted-tests` gives the answer); `run_all` true means no subset can be vouched for, with up to three `reasons` (`reasons_total` counts them all). `tests_to_run` (capped at ten, `total` and `truncated`) is in `repowise impacted-tests`' [run order](CLI_REFERENCE.md#repowise-impacted-tests-revspec), the `always_run_total` tests that run with every subset last; `why` gives each listed file's reason. With `run_all` the list is what to run first. `tests_to_run_kind` is `test_id` when a coverage node id is listed, else `test_file`. `basis` is `measured` when coverage decided a changed file, `inferred` when only the graph or a rule did, `none` when nothing named a test; `basis_by_file` names the evidence per changed file (`coverage`, `changed-test`, `call-graph`, `import-graph`, `conftest`, `full-run`, `no-tests-needed`, ...). With a coverage map, `map_present` is true and `line_coverage` buckets the scored files' changed lines, matched on the side of the diff the map was measured at (by file when it was measured at neither): `untested_changes`, `stale_test_candidates` (covered lines whose test file is absent from the diff), `covered`, `no_coverage_data`. Build the measured map with `coverage run --contexts=test` and `repowise coverage add`.
diff --git a/packages/cli/src/repowise/cli/commands/risk_cmd.py b/packages/cli/src/repowise/cli/commands/risk_cmd.py
index f651bb954..19c7aa286 100644
--- a/packages/cli/src/repowise/cli/commands/risk_cmd.py
+++ b/packages/cli/src/repowise/cli/commands/risk_cmd.py
@@ -234,13 +234,22 @@ def _render_card(name: str, card: dict) -> None:
console.print(f" [dim]unresolved: {reason}[/dim]")
return
trend = card.get("trend") or "unknown"
- console.print(
- f" [dim]hotspot {float(card.get('hotspot_score') or 0.0):.0%} ({trend}) · "
- f"{card.get('dependents_count', 0)} direct dependents · "
- f"{card.get('risk_type', 'unknown')} · owned "
- f"{_ta.owner_share(card.get('owner_pct'))} by "
- f"{escape(str(card.get('primary_owner') or 'unknown'))}[/dim]"
- )
+ if card.get("signal_tier") == "history":
+ # Nothing measured a hotspot, dependents or classification for this file.
+ last = str(card.get("last_commit_at") or "unknown")[:10]
+ console.print(
+ f" [dim]history only · {card.get('commit_count_total', 0)} commits ({trend}) · "
+ f"last {escape(last)} · owned {_ta.owner_share(card.get('owner_pct'))} by "
+ f"{escape(str(card.get('primary_owner') or 'unknown'))}[/dim]"
+ )
+ else:
+ console.print(
+ f" [dim]hotspot {float(card.get('hotspot_score') or 0.0):.0%} ({trend}) · "
+ f"{card.get('dependents_count', 0)} direct dependents · "
+ f"{card.get('risk_type', 'unknown')} · owned "
+ f"{_ta.owner_share(card.get('owner_pct'))} by "
+ f"{escape(str(card.get('primary_owner') or 'unknown'))}[/dim]"
+ )
# Kept in the projection, so each owes a line here: a key kept in the
# payload that no renderer prints is the second silent failure mode of a
# trim, and the one a projection test cannot see.
diff --git a/packages/core/src/repowise/core/analysis/change_health/models.py b/packages/core/src/repowise/core/analysis/change_health/models.py
index e709c81a0..2de2b8cbf 100644
--- a/packages/core/src/repowise/core/analysis/change_health/models.py
+++ b/packages/core/src/repowise/core/analysis/change_health/models.py
@@ -46,10 +46,18 @@ SEVERITY_RANK: dict[str, int] = {"low": 0, "medium": 1, "high": 2, "critical": 3
#: Documentation or configuration: skipped, but no code went unexamined.
NOT_CODE = "not_code"
+#: Tree-sitter queries and similar data: no detector reads them, yet they can
+#: change behaviour, so they are named apart from docs and config.
+NOT_ANALYZABLE_DATA = "not_analyzable_data"
+
+#: Skip reasons that leave no code unexamined.
+NO_CODE_GAP = frozenset({NOT_CODE, NOT_ANALYZABLE_DATA})
+
#: Why a changed file produced no comparison. Kept per file so a partial run
#: can never render as a clean one.
SkipReason = Literal[
"not_code",
+ "not_analyzable_data",
"unsupported_language",
"generated",
"binary",
@@ -171,7 +179,7 @@ class ChangeHealthDelta:
resolved_total: int = 0
unchanged_total: int = 0
uncertain_total: int = 0
- #: ``{path: reason}``. Any reason but ``not_code`` means the run was partial.
+ #: ``{path: reason}``. Any reason outside ``NO_CODE_GAP`` means the run was partial.
skipped: dict[str, str] = field(default_factory=dict)
timing_ms: float = 0.0
cache_hit: bool = False
@@ -188,7 +196,7 @@ class ChangeHealthDelta:
@property
def code_skipped(self) -> dict[str, str]:
"""Skipped files that may hold code, so their absence leaves a gap."""
- return {p: r for p, r in self.skipped.items() if r != NOT_CODE}
+ return {p: r for p, r in self.skipped.items() if r not in NO_CODE_GAP}
@property
def is_clean(self) -> bool:
diff --git a/packages/core/src/repowise/core/analysis/change_health/service.py b/packages/core/src/repowise/core/analysis/change_health/service.py
index 114ef58fb..cebaf3d9a 100644
--- a/packages/core/src/repowise/core/analysis/change_health/service.py
+++ b/packages/core/src/repowise/core/analysis/change_health/service.py
@@ -12,6 +12,7 @@ import threading
import time
from collections import OrderedDict
from dataclasses import dataclass
+from pathlib import PurePosixPath
from ...support_paths import is_doc_or_config_path
from ...test_paths import is_test_related_path
@@ -24,6 +25,8 @@ from .attribution import FindingAttributor, changed_symbols_for
from .identity import change_finding_id, finding_key, severity_rank
from .matcher import FindingMatcher, MatchedFinding
from .models import (
+ NO_CODE_GAP,
+ NOT_ANALYZABLE_DATA,
NOT_CODE,
AnalysisFingerprint,
ChangeFinding,
@@ -47,6 +50,12 @@ MAX_CHANGED_FILES = 300
#: dialect (HTML, schemas, shell) may carry logic, so it stays a gap.
_NOT_CODE_LANGUAGES = frozenset({"json", "yaml", "toml", "markdown"})
+#: Tree-sitter query files: patterns no health detector reads. ``.scm`` is also
+#: Scheme source, so only the tree-sitter layout (a ``queries`` directory)
+#: counts; any other ``.scm`` stays a gap.
+_QUERY_SUFFIX = ".scm"
+_QUERY_DIR = "queries"
+
#: Comparisons kept in the process cache.
_CACHE_CAPACITY = 32
@@ -218,19 +227,10 @@ class ChangeHealthDeltaService:
basis = "both_sides_analyzed"
if not eligible:
- if not changes:
- explanation = "This change touches no files."
- elif all(r == NOT_CODE for r in skipped.values()):
- explanation = (
- "Only documentation or configuration files changed; "
- "health analysis does not cover them."
- )
- else:
- explanation = "No changed file is health-analyzable, so nothing was compared."
delta = ChangeHealthDelta(
# Nothing was compared either way, so neither case is a clean bill.
status="unavailable",
- explanation=explanation,
+ explanation=_nothing_compared(changes, skipped),
base=base_id,
head=head_id,
comparison_basis=basis,
@@ -392,6 +392,20 @@ def _filter(changes: list[FileChange], request: DeltaRequest) -> list[FileChange
)
+def _nothing_compared(changes: list[FileChange], skipped: dict[str, str]) -> str:
+ """Why no changed file reached the analyzer, for the explanation."""
+ if not changes:
+ return "This change touches no files."
+ if all(r == NOT_CODE for r in skipped.values()):
+ return (
+ "Only documentation or configuration files changed; "
+ "health analysis does not cover them."
+ )
+ if all(r in NO_CODE_GAP for r in skipped.values()):
+ return "Only query or other data files changed; health analysis does not cover them."
+ return "No changed file is health-analyzable, so nothing was compared."
+
+
def _skip_reason(change: FileChange) -> str:
"""Why a changed path never reached the analyzer."""
if change.head_path is None:
@@ -400,6 +414,9 @@ def _skip_reason(change: FileChange) -> str:
if language in _NOT_CODE_LANGUAGES or is_doc_or_config_path(change.head_path):
return NOT_CODE
if language is None:
+ path = PurePosixPath(change.head_path)
+ if path.suffix.lower() == _QUERY_SUFFIX and _QUERY_DIR in path.parts[:-1]:
+ return NOT_ANALYZABLE_DATA
return "unsupported_language"
return "not_health_analyzable"
@@ -413,7 +430,7 @@ def _status_for(
f"{'file' if scope.changed == 1 else 'files'} could be analysed."
)
# Docs, config and data hold no code, so skipping them leaves nothing unexamined.
- code_skipped = sum(1 for r in skipped.values() if r != NOT_CODE)
+ code_skipped = sum(1 for r in skipped.values() if r not in NO_CODE_GAP)
if code_skipped or scope.failed:
return (
"partial",
diff --git a/packages/core/src/repowise/core/analysis/decisions/rationale_comments.py b/packages/core/src/repowise/core/analysis/decisions/rationale_comments.py
index 4326d0466..aa95bddaf 100644
--- a/packages/core/src/repowise/core/analysis/decisions/rationale_comments.py
+++ b/packages/core/src/repowise/core/analysis/decisions/rationale_comments.py
@@ -183,6 +183,8 @@ LINE_PREFIXES: dict[str, tuple[str, ...]] = {
"lua": ("--",),
"hs": ("--",),
"elm": ("--",),
+ # Tree-sitter query files (and Scheme). Longest prefix first: ";; x" is common.
+ "scm": (";;", ";"),
}
# Extensions that use C-style /* ... */ block comments.
@@ -312,7 +314,7 @@ def _strip_comment_open(stripped: str, prefixes: tuple[str, ...]) -> str | None:
# A line that is only a divider rule (``# -------`` / ``// =====`` / ``***``).
# These delimit comment runs and must never be glued into the rationale text.
-_SEPARATOR_CHARS = frozenset("-=*~_#/+.<> ")
+_SEPARATOR_CHARS = frozenset("-=*~_#/+.<>; ")
def _is_separator(body: str) -> bool:
diff --git a/packages/core/src/repowise/core/analysis/review_directive.py b/packages/core/src/repowise/core/analysis/review_directive.py
index ed4399272..6b3200e68 100644
--- a/packages/core/src/repowise/core/analysis/review_directive.py
+++ b/packages/core/src/repowise/core/analysis/review_directive.py
@@ -15,7 +15,7 @@ import hashlib
from dataclasses import dataclass
from typing import Literal
-from .change_health.models import ChangeHealthDelta
+from .change_health.models import NOT_ANALYZABLE_DATA, ChangeHealthDelta
from .health.scoring import ADVISORY_DIMENSION
#: How much of a lane's evidence was actually computed.
@@ -120,14 +120,21 @@ def _plural(word: str, count: int) -> str:
return word if count == 1 else f"{word}s"
-def _verdict(delta: ChangeHealthDelta) -> tuple[ReviewStatus, str, EvidenceState]:
- if delta.status == "unavailable" and delta.skipped and not delta.code_skipped:
+def _no_code_headline(delta: ChangeHealthDelta) -> str:
+ if NOT_ANALYZABLE_DATA in delta.skipped.values():
return (
- "clear_in_analyzed_scope",
- "No code changed: only documentation or configuration files; "
- "health analysis does not cover them.",
- "available",
+ "No code changed: only query or other data files, which health analysis "
+ "does not cover. They can still change behaviour, so test selection still applies."
)
+ return (
+ "No code changed: only documentation or configuration files; "
+ "health analysis does not cover them."
+ )
+
+
+def _verdict(delta: ChangeHealthDelta) -> tuple[ReviewStatus, str, EvidenceState]:
+ if delta.status == "unavailable" and delta.skipped and not delta.code_skipped:
+ return "clear_in_analyzed_scope", _no_code_headline(delta), "available"
if delta.status in _NOT_COMPARED:
return "unknown", f"Change health could not be compared: {delta.explanation}", "unavailable"
if delta.status in _UNTRUSTWORTHY:
diff --git a/packages/server/src/repowise/server/mcp_server/tool_risk/assessment.py b/packages/server/src/repowise/server/mcp_server/tool_risk/assessment.py
index 6bc583134..86c7e01ba 100644
--- a/packages/server/src/repowise/server/mcp_server/tool_risk/assessment.py
+++ b/packages/server/src/repowise/server/mcp_server/tool_risk/assessment.py
@@ -434,6 +434,47 @@ async def _unresolved_reason(
return {"unresolved_reason": "no_such_path", **hint}
+def _history_tier_card(target: str, meta: Any, security_signals: list[dict]) -> dict:
+ """The card for a file indexed at the history tier (``meta.history_only``).
+
+ A non-code file keeps commit counts, windows and authorship, but no churn
+ percentile, hotspot flag, co-change partners or fix rollup, so the scored
+ card would print zeros nothing measured. This one carries only what was.
+ """
+ commits = meta.commit_count_total or 0
+ last = meta.last_commit_at.isoformat() if meta.last_commit_at else None
+ owner = meta.primary_owner_name or "unknown"
+ pct = meta.primary_owner_commit_pct or 0.0
+ return {
+ "target": target,
+ "signal_tier": "history",
+ "signal_tier_reason": (
+ "not a parsed code file: only commit history is indexed, so hotspot, "
+ "co-change and risk classification do not apply"
+ ),
+ "commit_count_total": commits,
+ "commit_count_90d": meta.commit_count_90d or 0,
+ "last_commit_at": last,
+ "trend": _compute_trend(meta),
+ "primary_owner": owner,
+ "owner_pct": pct,
+ "contributor_count": meta.contributor_count or 0,
+ "security_signals": security_signals,
+ # Workspace enrichment adds its rows here; co-change covers code files only.
+ "relationship_analysis": {
+ "co_change": {
+ "status": "unavailable",
+ "scope": "git_history",
+ "reason": "history-tier file",
+ }
+ },
+ "risk_summary": (
+ f"{target}: history only, {commits} commits, last {last[:10] if last else 'unknown'}, "
+ f"{_owner_clause(owner, pct, None)}"
+ ),
+ }
+
+
async def _assess_one_target(
session: AsyncSession,
repository: Repository,
@@ -588,6 +629,10 @@ async def _assess_one_target(
),
}
+ if getattr(meta, "history_only", False) is True:
+ security = await _get_security_signals(session, repo_id, lookup_path)
+ return _history_tier_card(target, meta, security)
+
if meta is None:
result_data["hotspot_score"] = 0.0
result_data["is_hotspot"] = False
diff --git a/packages/server/src/repowise/server/mcp_server/tool_risk/directives.py b/packages/server/src/repowise/server/mcp_server/tool_risk/directives.py
index e7cbe87b8..509deaba6 100644
--- a/packages/server/src/repowise/server/mcp_server/tool_risk/directives.py
+++ b/packages/server/src/repowise/server/mcp_server/tool_risk/directives.py
@@ -629,8 +629,8 @@ def _build_pr_directive(
# response carries an expandable [repowise#<ref>] marker for it.
for r in response["targets"].values():
# The counts below would be the structural zeros get_risk stopped
- # emitting for a card that resolved nothing.
- if r.get("resolved") is False:
+ # emitting for a card that resolved nothing or holds history only.
+ if r.get("resolved") is False or r.get("signal_tier") == "history":
continue
partners = r.get("co_change_partners") or []
if len(partners) > 3:
diff --git a/tests/unit/analysis/test_decision_rationale_comments.py b/tests/unit/analysis/test_decision_rationale_comments.py
index 7f7751528..7f7474677 100644
--- a/tests/unit/analysis/test_decision_rationale_comments.py
+++ b/tests/unit/analysis/test_decision_rationale_comments.py
@@ -62,6 +62,18 @@ def test_extract_comment_blocks_handles_c_style_and_returns_empty_for_unknown():
assert extract_comment_blocks("plain prose because yes", "txt") == []
+def test_tree_sitter_query_comments_are_mined():
+ scm = (
+ ";;;;;;;;;;;;;;;;\n"
+ ";; The receiver stays uncaptured because a qualified super call\n"
+ "; binds to the parent class, not the field.\n"
+ "(method_invocation name: (identifier) @call.target) @call.site\n"
+ )
+ blocks = extract_comment_blocks(scm, "scm")
+ assert blocks[0].text.startswith("The receiver stays uncaptured")
+ assert harvest_file_rationale(scm, "scm")
+
+
# ---------------------------------------------------------------------------
# Guardrails
# ---------------------------------------------------------------------------
diff --git a/tests/unit/analysis/test_review_directive.py b/tests/unit/analysis/test_review_directive.py
index 4a918b640..330abe856 100644
--- a/tests/unit/analysis/test_review_directive.py
+++ b/tests/unit/analysis/test_review_directive.py
@@ -137,6 +137,16 @@ def test_an_unsupported_code_file_still_leaves_the_verdict_unknown():
assert [a.targets for a in directive.actions] == [("lib.ex",)]
+def test_a_query_only_change_is_clear_with_its_own_headline():
+ delta = _delta(status="unavailable", explanation="Only query files changed")
+ delta.skipped = {"queries/java.scm": "not_analyzable_data", "README.md": "not_code"}
+ directive = review_directive(delta)
+ assert directive.status == "clear_in_analyzed_scope"
+ assert "query" in directive.headline
+ assert "documentation or configuration" not in directive.headline
+ assert not directive.actions
+
+
# ---------------------------------------------------------------------------
# Actions carry targets, not rendered strings
# ---------------------------------------------------------------------------
diff --git a/tests/unit/change_health/test_comparison.py b/tests/unit/change_health/test_comparison.py
index 7064febdb..e162472c2 100644
--- a/tests/unit/change_health/test_comparison.py
+++ b/tests/unit/change_health/test_comparison.py
@@ -598,3 +598,39 @@ def test_the_analyzer_fingerprint_is_reported_on_both_sides(make_repo):
assert delta.fingerprint.analyzer_version > 0
assert delta.fingerprint.performance_model_version > 0
assert delta.comparison_basis == "both_sides_analyzed"
+
+
+def test_a_query_file_change_is_data_not_a_gap(make_repo):
+ repo = make_repo()
+ repo.commit("seed", {"queries/java.scm": "(identifier) @name\n"})
+ repo.commit("edit", {"queries/java.scm": "(type_identifier) @name\n"})
+
+ delta = compare(repo)
+
+ assert delta.status == "unavailable"
+ assert delta.skipped == {"queries/java.scm": "not_analyzable_data"}
+ assert delta.code_skipped == {}
+ assert "query" in delta.explanation
+ assert not delta.is_clean
+
+
+def test_a_query_file_beside_code_does_not_make_the_run_partial(make_repo):
+ repo = make_repo()
+ repo.commit("seed", {"app.py": "x = 1\n", "queries/java.scm": "(identifier) @a\n"})
+ repo.commit("edit", {"app.py": "x = 2\n", "queries/java.scm": "(identifier) @b\n"})
+
+ delta = compare(repo)
+
+ assert delta.status == "available"
+ assert delta.code_skipped == {}
+
+
+def test_a_scheme_file_outside_queries_is_still_a_gap(make_repo):
+ repo = make_repo()
+ repo.commit("seed", {"src/main.scm": "(define (f x) x)\n"})
+ repo.commit("edit", {"src/main.scm": "(define (f x) (+ x 1))\n"})
+
+ delta = compare(repo)
+
+ assert delta.skipped == {"src/main.scm": "unsupported_language"}
+ assert delta.code_skipped == {"src/main.scm": "unsupported_language"}
diff --git a/tests/unit/cli/test_risk_cmd_contract.py b/tests/unit/cli/test_risk_cmd_contract.py
index ef8b5b872..1d96996d5 100644
--- a/tests/unit/cli/test_risk_cmd_contract.py
+++ b/tests/unit/cli/test_risk_cmd_contract.py
@@ -204,6 +204,32 @@ def test_the_projection_survives_a_payload_with_no_targets_key() -> None:
assert project_risk({})["targets"] == {}
+def test_a_history_tier_card_prints_its_history_not_hotspot_zeros() -> None:
+ from unittest.mock import patch
+
+ from rich.console import Console
+
+ from repowise.cli.commands import risk_cmd
+
+ card = {
+ "signal_tier": "history",
+ "commit_count_total": 11,
+ "last_commit_at": "2026-03-01T00:00:00",
+ "trend": "stable",
+ "primary_owner": "Carol",
+ "owner_pct": 0.5,
+ }
+ projected = project_risk({"targets": {"queries/java.scm": card}})["targets"]["queries/java.scm"]
+ assert projected["signal_tier"] == "history" and projected["commit_count_total"] == 11
+ rec = Console(record=True, width=300)
+ with patch.object(risk_cmd, "console", rec):
+ risk_cmd._render_card("queries/java.scm", projected)
+ out = rec.export_text()
+ assert "history only · 11 commits (stable) · last 2026-03-01" in out
+ assert "Carol" in out
+ assert "hotspot" not in out and "direct dependents" not in out
+
+
# ---------------------------------------------------------------------------
# Errors a caller scripts against
# ---------------------------------------------------------------------------
diff --git a/tests/unit/server/mcp/test_risk.py b/tests/unit/server/mcp/test_risk.py
index c9e69e12d..6eb023dd5 100644
--- a/tests/unit/server/mcp/test_risk.py
+++ b/tests/unit/server/mcp/test_risk.py
@@ -1159,3 +1159,48 @@ async def test_get_risk_pr_mode_puts_docs_and_config_cards_last(setup_mcp):
files = ["CHANGES.rst", "setup.cfg", "src/auth/service.py"]
result = await get_risk(files, changed_files=files)
assert next(iter(result["targets"])) == "src/auth/service.py"
+
+
[email protected]("path", ["queries/java.scm", "docs/guide.md"])
+async def test_get_risk_history_tier_file_omits_unmeasured_signals(setup_mcp, session, path):
+ """A history-only row carries commits and owner, never zeros posing as measurements."""
+ from datetime import UTC, datetime
+
+ from repowise.core.persistence.models import GitMetadata
+ from repowise.server.mcp_server import get_risk
+
+ session.add(
+ GitMetadata(
+ id=f"gm-history-{path}",
+ repository_id=setup_mcp,
+ file_path=path,
+ history_only=True,
+ commit_count_total=11,
+ commit_count_90d=4,
+ commit_count_30d=1,
+ last_commit_at=datetime(2026, 3, 1, tzinfo=UTC),
+ primary_owner_name="Carol",
+ primary_owner_commit_pct=0.5,
+ top_authors_json=json.dumps([{"name": "Carol", "count": 6}]),
+ )
+ )
+ await session.flush()
+
+ for result in (
+ await get_risk([path], include=["graph", "churn"]),
+ await get_risk([path], changed_files=[path]),
+ ):
+ card = result["targets"][path]
+ assert card["signal_tier"] == "history"
+ assert card["commit_count_total"] == 11
+ assert card["last_commit_at"].startswith("2026-03-01")
+ assert card["primary_owner"] == "Carol"
+ for key in ("hotspot_score", "is_hotspot", "co_change_partners", "risk_type"):
+ assert key not in card
+ assert "co_change_partners_emitted" not in card
+ assert "hotspot" not in card["risk_summary"]
+ # Detailed owner metrics stay behind include=["owners"], as on a scored card.
+ assert "owner_pct" not in card and "contributor_count" not in card
+ owners = (await get_risk([path], include=["owners"]))["targets"][path]
+ assert owners["owner_pct"] == 0.5