feat(model): first repository measurement, and the four things it found

contributor-scan 2.2. VAmPI at a pinned commit, scanned whole rather than
as excerpts, with benchmarks/repos/measure.py recording build time, query
time, arbitration time, peak memory, cost, unjudged regions, and whether
each known vulnerability was found and at which rung.

The corpus could not have found any of this. Every defect below needs a
repository to become visible, and three of them were invisible behind a
passing test suite.

**The batch path was unreachable.** query_batch was implemented, tested and
committed while CpgResult exposed only `run`, so the driver's capability
check found nothing and fell back to one JVM per region-family: the first
attempt at this measurement ran eighteen minutes on 520 lines, issuing
`--param function=` one at a time. Every test called query_batch directly,
so none of them noticed. CpgResult now carries run_batch and a test asserts
the capability is reachable through the object the driver is handed.
The earlier 275s->106s improvement attributed to batching came from region
deduplication alone.

**The budget stopped free work.** It bounds model calls, but was enforced
by breaking the region loop -- spending a paid limit to stop unpaid CPG
arbitration. 6 of 22 regions were examined and VAmPI's BOLA, which the
graph entails for nothing, was never reached. It now withholds the judge
only. 22 of 22 regions arbitrated, 17 counted as regions_unasked (silent
because unaffordable is not silent because decided), and the BOLA is
entailed by the model layer.

**A region with no function asked about the whole repository.** Sending
function="" matched every hit in the tree and the driver attributed that
answer to whoever asked: one `host='0.0.0.0'` in app.py became eight
identical entailed findings in eight files that do not contain it. Queries
now carry the region's file. Entailed findings 14 -> 7, query time
70.9s -> 40.9s. The single-request form stays unscoped so the committed
pair measurements remain reproducible.

**The JVM had no heap bound.** Joern claimed a quarter of physical RAM and
runs were killed under memory pressure at different points, so the numbers
described the host rather than the engine. -J-Xmx2048m by default,
OPENULTRASAST_CPG_HEAP_MB to raise it.

Also: arbitrated findings carried `function:line` where the report layer
read `path:line`, so every entailed finding reached a contributor with a
function name where its file belonged; and the witness line was parsed by
splitting on a comma, which produced sites like
`config.py:17): permissive literal '0.0.0.0':?`.

Reading: GO on one CPG per repository -- 6.2s to build, 177s end to end,
869MB, $0.0067, nothing infeasible. NO-GO on region-scoped taint as the
only shape: VAmPI's SQL injection runs from a handler parameter into a
function in another module, and a per-function region sees neither end.
That is the ordinary shape of a real injection bug, so it is recorded as
task 2.4 along with three smaller defects this run exposed.

Gates byte-identical.

Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_01T7XD7x5KonMdmGRkp63EQG

tests/test_model_pipeline.py

@@ -79,6 +79,22 @@ def test_an_entailed_site_becomes_a_finding_with_no_model_call() -> None:
     assert client.calls == 0
 
 
+def test_an_arbitrated_finding_names_its_file_the_way_a_candidate_does() -> None:
+    """One consumer parses both shapes, so both must be ``path:line:function``.
+
+    An arbitrated finding was ``function:line``, which the report layer read as ``path:line``: every entailed
+    finding reached a contributor with a function name where its file belonged and no line at all. A whole
+    repository scan is what made it visible -- `model:access_control:get_all_books:?`, path `get_all_books`.
+    """
+    from openultrasast.model.pipeline import scan_region
+
+    findings = scan_region(_cpg(ENTAILING), _spec(), path="models/user_model.py", function="get_user", client=_Client(), model="m")
+
+    path, _, rest = findings[0].site.partition(":")
+    assert path == "models/user_model.py", "the file, not the function"
+    assert rest.endswith(":get_user")
+
+
 def test_a_sanitized_flow_the_judge_confirms_becomes_corroborated() -> None:
     from openultrasast.model.ladder import Rung
     from openultrasast.model.pipeline import scan_region
@@ -244,3 +260,16 @@ def test_the_dominance_question_describes_the_evidence_the_model_actually_has()
     assert "siblings have" not in q, "the corroborated case has no guarded siblings to appeal to"
     assert "sibling handlers are guarded" not in q
     assert "guard" in q.lower() and "public" in q.lower(), "it must still ask whether the operation needs a guard"
+
+
+def test_a_witness_line_is_read_whatever_follows_the_number() -> None:
+    """The delimiter after the line number differs by family: `(line 17,` in one, `(line 17):` in another.
+
+    Splitting on the comma produced sites like `config.py:17): permissive literal '0.0.0.0':?` -- a line
+    number no editor can open and no consumer can compare.
+    """
+    from openultrasast.model.pipeline import _site
+
+    assert _site("config.py", "", "permissive literal '0.0.0.0' (line 17): permissive") == "config.py:17:?"
+    assert _site("a.py", "run", "flow reaches os.system (line 9, via x)") == "a.py:9:run"
+    assert _site("a.py", "run", "no line at all") == "a.py:?:run"