open/portfolio-optimiser
1
0
Fork 0
Code Issues Releases 3 Activity

test(v1-gate): the v1 gate, written RED

Browse source 
One command says how far the repo is from v1, row by row, with an exit
code: rounds with a real domain expert 0/3, traced measurable change 0/3,
feedback types with a way in and an action 3/8 (1, 3, 7), round 3 report
kept - none, MAF points with a green type pointer 0/8 (list not approved),
validated without the approach's own declaration 10/10 in stress round 6,
and `named` 1/20 as a diagnosis that never moves the exit code.

The gate defines the contract (a fixed rounds directory, gitignored by
default), not the generator. Rows 3 and 6 run named tests with
--runxfail; the red probes are xfail(strict=True) so the suite stays green
while the gap is real. No product code changed.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
Kjell Tore Guttormsen 2026-09-17 15:31:48 +02:00
commit 83c94e4fb6
7 changed files with 1397 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

1
src/portfolio_optimiser/evals/__init__.py Normal file
View file

@ -0,0 +1 @@
"""Deterministic evaluation gates (no model calls)."""

87
src/portfolio_optimiser/evals/v1_gate.json Normal file
View file

@ -0,0 +1,87 @@
{
"rounds_required": 3,
"keep_threshold": 0.8,
"ai_authored": ["docs/ekspert-svar.md"],
"feedback_types": {
"1": {
"label": "feil - rett en feil i forslaget",
"evidence": [
"tests/test_proposal_review_loop_loadbearing.py::test_t13_the_flag_answers_the_review_from_a_real_argv_and_the_answer_is_used"
]
},
"2": {
"label": "bort - ta en retning bort",
"evidence": ["tests/test_v1_probes.py::test_type_2_remove_a_direction_has_a_typed_door"]
},
"3": {
"label": "vinklinger - nye vinklinger",
"evidence": ["tests/test_mandate_cli.py::test_run_settles_against_the_mandate_afterwards"]
},
"4": {
"label": "lette paa krav",
"evidence": ["tests/test_v1_probes.py::test_type_4_relax_a_requirement_has_a_door"]
},
"5": {
"label": "konseptgrafer",
"evidence": ["tests/test_v1_probes.py::test_type_5_edit_the_concept_graph_has_a_door"]
},
"6": {
"label": "skills per analyse",
"evidence": ["tests/test_v1_probes.py::test_type_6_skills_per_analysis_has_a_door"]
},
"7": {
"label": "MCP - verktoey i debatten",
"evidence": [
"tests/test_mcp_run_loadbearing.py::test_configured_server_becomes_a_tool_the_agents_have",
"tests/test_b4_mcp_call_trace_loadbearing.py::test_a_called_mcp_tool_is_recorded_in_provenance"
]
},
"8": {
"label": "inline kontekst",
"evidence": ["tests/test_v1_probes.py::test_type_8_inline_context_has_a_door"]
}
},
"row6_evidence": [
"tests/test_v1_probes.py::test_row6_an_approach_that_declared_nothing_cannot_be_validated",
"tests/test_v1_probes.py::test_row6_a_run_level_declaration_does_not_stand_in_for_the_approach"
],
"maf_points": {
"approved": false,
"points": [
{"u_id": "U13", "construct": "MagenticPlanReviewResponse", "package": "agent_framework",
"callsite": {"module": "explore.py", "scope": "_drive"}, "types": [1, 2]},
{"u_id": "U9", "construct": "ContextProvider", "package": "agent_framework",
"callsite": {"module": "verdicts.py", "scope": "ExpeLContextProvider"}, "types": [1, 8]},
{"u_id": "U12", "construct": "FileCheckpointStorage", "package": "agent_framework",
"callsite": {"module": "explore.py", "scope": "checkpoint_storage"}, "types": [1]},
{"u_id": "U4", "construct": "MagenticBuilder", "package": "agent_framework",
"callsite": {"module": "explore.py", "scope": "fresh_exploration_workflow"}, "types": [3]},
{"u_id": "U7", "construct": "tool", "package": "agent_framework",
"callsite": {"module": "explore.py", "scope": "navigator_tools"}, "types": [4]},
{"u_id": "U11", "construct": "TextSpanRegion", "package": "agent_framework",
"callsite": {"module": "provenance.py", "scope": "ProvenanceStamp"}, "types": [5]},
{"u_id": "U5", "construct": "SkillsProvider", "package": "agent_framework",
"callsite": null, "types": [6]},
{"u_id": "U6", "construct": "MCPStdioTool", "package": "agent_framework",
"callsite": {"module": "mcp_tools.py", "scope": "build_mcp_tools"}, "types": [7]}
]
},
"stress_evidence": {
"label": "stressrunde 6",
"root": "scratchpad/p22-stress",
"runs": [
{"context": "contexts/gate-nordvik-2027", "outbox": "gate-nordvik-2027",
"run_id": "gate-nordvik-2027-08"},
{"context": "contexts/tunnel-hauglia-2027", "outbox": "tunnel-hauglia-2027",
"run_id": "tunnel-hauglia-2027-08"},
{"context": "contexts/fv412-dekkefornyelse-2027", "outbox": "fv412-dekkefornyelse-2027",
"run_id": "fv412-dekkefornyelse-2027-08"},
{"context": "contexts/kontrakt-sorasen-2027", "outbox": "kontrakt-sorasen-2027",
"run_id": "kontrakt-sorasen-2027-08"},
{"context": "contexts/dekke-og-kontrakt-lindaas-2027", "outbox": "lindaas",
"run_id": "lindaas-04-vegnormal-n200-2024", "bundle": "vegnormal-n200-2024"},
{"context": "contexts/dekke-og-kontrakt-lindaas-2027", "outbox": "lindaas",
"run_id": "lindaas-04-vegnormal-r761-2025", "bundle": "vegnormal-r761-2025"}
]
}
}

722
src/portfolio_optimiser/evals/v1_gate.py Normal file
View file

@ -0,0 +1,722 @@
"""The v1 gate: how far portfolio-optimiser is from v1, row by row, with an exit code.
One command, deterministic, offline: no model call, no network, no quota. Exit 0 only when every
FAILING row (1-6) is green; exit 1 otherwise; exit 2 on wrong usage. Row 7 is a diagnosis and never
moves the exit code.
The gate defines the CONTRACT a later capability must deliver into, not the generator. It reads a
rounds directory of a fixed shape (see ``--help``); nothing in the product writes it yet. Every
denominator comes from a source outside the thing being measured: the number of rounds is the
operator's choice (3), the feedback types are the eight named in ``v1_gate.json``, the MAF points
are a listed, approved-or-not set, and rows 6-7 read the stress round's own artefacts.
Rows 3 and 6 run NAMED tests (``v1_gate.json``) in a child pytest with ``--runxfail``: a type
counts only when every test registered for it passes, so a missing surface is a red test, never a
missing one.
"""
from __future__ import annotations
import argparse
import ast
import json
import os
import subprocess
import sys
import tempfile
import xml.etree.ElementTree as ET
from collections import Counter
from collections.abc import Callable, Iterable, Mapping, Sequence
from dataclasses import asdict, dataclass, field
from pathlib import Path
from typing import Any
_DATA = Path(__file__).with_name("v1_gate.json")
_PACKAGE_SRC = Path(__file__).resolve().parents[1]
_REPO_ROOT = Path(__file__).resolve().parents[3]
#: Gitignored by default: the remote is public, and a domain expert's feedback must never reach it.
DEFAULT_ROUNDS_DIR = "v1-rounds"
#: A line of an AI-authored document shorter than this is too generic to identify its origin.
_AI_LINE_MIN = 30
GREEN = "GRØNN"
RED = "RØD"
DIAGNOSIS = "DIAGNOSE"
ROUNDS_CONTRACT = """\
Rundekatalogen (--rounds-dir) har fast form:
<rounds-dir>/0/outcome.json grunnkjøringen (runde 0) som runde 1 måles mot
<rounds-dir>/<n>/report.md rapporten slik fagpersonen fikk den (n = 1, 2, 3)
<rounds-dir>/<n>/feedback.json fagpersonens tilbakemelding
<rounds-dir>/<n>/outcome.json utfallet rapporten er bygget fra
<rounds-dir>/3/report.kept.md runde 3-rapporten slik fagpersonen BEHOLDT den
feedback.json:
{"author": "<fagpersonen>",
"items": [{"id": "<unik id>", "type": <1-8>, "text": "<tilbakemeldingen>"}]}
outcome.json:
{"run_id": "<kjøringen>",
"approaches": [{"id": "<tilnærming>", "validated": true|false,
"stage": "<avvisningsstadium, tom når validert>",
"validated_nok": <tall eller null>,
"feedback_ids": ["<id-er fra feedback.json som forklarer raden>"]}],
"removed": [{"id": "<tilnærming fjernet siden forrige runde>", "feedback_ids": [...]}]}
En runde har målbar endring når den skiller seg fra forrige på minst én av (a) settet av
tilnærmings-id-er, (b) hvilke som er validert, (c) avvisningsstadium, (d) validert NOK — OG minst
én endret rad bærer en feedback-id gitt i DENNE runden. Tekst tatt fra et AI-forfattet dokument
(docs/ekspert-svar.md) teller aldri som fagperson-tilbakemelding.
"""
@dataclass(frozen=True)
class Row:
key: str
title: str
k: int | None
n: int | None
status: str
reason: str
failing: bool = True
exceptions: tuple[str, ...] = ()
diagnostics: tuple[str, ...] = ()
def line(self) -> str:
k = "–" if self.k is None else str(self.k)
n = "–" if self.n is None else str(self.n)
return f"{self.title} | {k} av {n} | {self.status} | {self.reason}"
def load_config(path: Path = _DATA) -> dict[str, Any]:
data: dict[str, Any] = json.loads(path.read_text(encoding="utf-8"))
return data
# ---------------------------------------------------------------------------------------------
# Rows 1 and 2 — rounds
# ---------------------------------------------------------------------------------------------
def _norm(text: str) -> str:
return " ".join(text.split())
def ai_authored_lines(repo_root: Path, docs: Sequence[str]) -> tuple[str, str] | None:
"""The normalised full text and the joined long lines of every AI-authored document, or
``None`` when one of them cannot be read — the guard then cannot run, and a round it cannot
check is never counted."""
texts: list[str] = []
for rel in docs:
path = repo_root / rel
if not path.is_file():
return None
texts.append(path.read_text(encoding="utf-8"))
lines = [
_norm(line.strip().lstrip(">*-#|` ").strip())
for text in texts
for line in text.splitlines()
]
return _norm("\n".join(texts)), "\n".join(x for x in lines if len(x) >= _AI_LINE_MIN)
def _is_ai_text(text: str, ai: tuple[str, str]) -> bool:
whole, lines = ai
item = _norm(text)
if len(item) >= _AI_LINE_MIN and item in whole:
return True
return any(line in item for line in lines.splitlines() if line)
def read_feedback(round_dir: Path, ai: tuple[str, str] | None) -> tuple[set[str], str]:
"""The ids of a round's feedback items, and ``""`` — or an empty set and the reason."""
path = round_dir / "feedback.json"
if not path.is_file():
others = sorted(p.name for p in round_dir.glob("feedback.*")) if round_dir.is_dir() else []
extra = f" (fant {', '.join(others)}; kontrakten er feedback.json)" if others else ""
return set(), f"runde {round_dir.name}: feedback.json mangler{extra}"
try:
data = json.loads(path.read_text(encoding="utf-8"))
author = str(data["author"]).strip()
items = list(data["items"])
except (ValueError, KeyError, TypeError) as exc:
return set(), f"runde {round_dir.name}: feedback.json uleselig ({exc!r})"
if not author:
return set(), f"runde {round_dir.name}: feedback.json navngir ingen fagperson"
if ai is None:
return set(), f"runde {round_dir.name}: AI-vakten kunne ikke lese sine kilder"
ids: set[str] = set()
for item in items:
try:
item_id, item_type, text = str(item["id"]), int(item["type"]), str(item["text"])
except (KeyError, TypeError, ValueError):
return set(), f"runde {round_dir.name}: et feedback-punkt mangler id/type/text"
if not item_id or not text.strip() or not 1 <= item_type <= 8:
return set(), f"runde {round_dir.name}: punkt {item_id!r} er tomt eller har ukjent type"
if _is_ai_text(text, ai):
return set(), f"runde {round_dir.name}: punkt {item_id!r} er AI-forfattet tekst"
ids.add(item_id)
if not ids:
return set(), f"runde {round_dir.name}: feedback.json har ingen punkter"
return ids, ""
def score_rounds(rounds_dir: Path, required: int, ai: tuple[str, str] | None) -> Row:
exceptions: list[str] = []
k = 0
if not rounds_dir.is_dir():
exceptions.append(f"{rounds_dir} finnes ikke")
else:
for n in range(1, required + 1):
ids, why = read_feedback(rounds_dir / str(n), ai)
if ids:
k += 1
else:
exceptions.append(why)
status = GREEN if k == required else RED
reason = "alle runder har fagperson-tilbakemelding" if k == required else exceptions[0]
return Row(
"rounds",
"1 runder med ekte fagperson",
k,
required,
status,
reason,
exceptions=tuple(exceptions),
)
def _read_outcome(path: Path) -> tuple[dict[str, dict[str, Any]], dict[str, set[str]], str, str]:
"""Rows by approach id, removed ids with their feedback ids, the run id, and ``""`` or why."""
try:
data = json.loads(path.read_text(encoding="utf-8"))
rows = {str(a["id"]): a for a in data["approaches"]}
removed = {
str(r["id"]): set(map(str, r.get("feedback_ids", ()))) for r in data.get("removed", ())
}
return rows, removed, str(data.get("run_id", "")), ""
except FileNotFoundError:
return {}, {}, "", f"{path} mangler"
except (ValueError, KeyError, TypeError) as exc:
return {}, {}, "", f"{path} uleselig ({exc!r})"
def _row_key(row: Mapping[str, Any]) -> tuple[bool, str, Any]:
return bool(row.get("validated")), str(row.get("stage") or ""), row.get("validated_nok")
def round_changed(before: Path, after: Path, feedback_ids: set[str]) -> tuple[bool, str]:
"""Whether round ``after`` changed measurably against ``before`` AND the change is traced to
feedback given in this round. The second half is what keeps model noise out."""
prev, _, _, why = _read_outcome(before)
if why:
return False, why
cur, removed, _, why = _read_outcome(after)
if why:
return False, why
changed: dict[str, set[str]] = {}
for aid, row in cur.items():
if aid not in prev or _row_key(prev[aid]) != _row_key(row):
changed[aid] = set(map(str, row.get("feedback_ids", ())))
for aid in prev.keys() - cur.keys():
changed[aid] = removed.get(aid, set())
if not changed:
return False, "ingen endring i (a)-(d)"
traced = sorted(aid for aid, ids in changed.items() if ids & feedback_ids)
if not traced:
return False, f"{len(changed)} rad(er) endret, ingen sporet til rundens feedback-id-er"
return True, f"{len(changed)} rad(er) endret, sporet: {', '.join(traced)}"
def score_changes(rounds_dir: Path, required: int, ai: tuple[str, str] | None) -> Row:
exceptions: list[str] = []
k = 0
_, _, base_run, base_why = _read_outcome(rounds_dir / "0" / "outcome.json")
base = f"runde 0 = {rounds_dir / '0' / 'outcome.json'}"
base += f" ({base_run})" if base_run else (f" — {base_why}" if base_why else "")
for n in range(1, required + 1):
ids, why = read_feedback(rounds_dir / str(n), ai)
if not ids:
exceptions.append(why)
continue
ok, detail = round_changed(
rounds_dir / str(n - 1) / "outcome.json", rounds_dir / str(n) / "outcome.json", ids
)
if ok:
k += 1
else:
exceptions.append(f"runde {n}: {detail}")
status = GREEN if k == required else RED
return Row(
"changes",
"2 runder med målbar endring",
k,
required,
status,
base,
exceptions=tuple(exceptions),
)
# ---------------------------------------------------------------------------------------------
# Row 3 and the row 6 probes — named tests run in a child
# ---------------------------------------------------------------------------------------------
ProbeRunner = Callable[[Sequence[str]], Mapping[str, str]]
def run_probes(nodeids: Sequence[str], repo_root: Path = _REPO_ROOT) -> dict[str, str]:
"""``nodeid -> "passed" | "failed" | "skipped" | "missing"``, from a child pytest run with
``--runxfail`` so a known gap shows as the failure it is."""
outcomes = {n: "missing" for n in nodeids}
if not (repo_root / "tests").is_dir():
return outcomes
env = {**os.environ, "PYTHONDONTWRITEBYTECODE": "1"}
base = [sys.executable, "-m", "pytest", "-p", "no:cacheprovider", "-q"]
files = sorted(
{n.split("::", 1)[0] for n in nodeids if (repo_root / n.split("::")[0]).is_file()}
)
if not files:
return outcomes
collect = subprocess.run(
[*base, "--collect-only", *files], cwd=repo_root, env=env, capture_output=True, text=True
)
collected = {line.strip() for line in collect.stdout.splitlines() if "::" in line}
present = [n for n in nodeids if n in collected]
if not present:
return outcomes
with tempfile.TemporaryDirectory() as tmp:
junit = Path(tmp) / "junit.xml"
subprocess.run(
[*base, "--runxfail", f"--junitxml={junit}", *present],
cwd=repo_root,
env=env,
capture_output=True,
text=True,
)
if not junit.is_file():
return {n: ("failed" if n in present else "missing") for n in nodeids}
for case in ET.parse(junit).getroot().iter("testcase"):
nodeid = case.get("classname", "").replace(".", "/") + ".py::" + case.get("name", "")
if nodeid not in outcomes:
continue
if case.find("failure") is not None or case.find("error") is not None:
outcomes[nodeid] = "failed"
elif case.find("skipped") is not None:
outcomes[nodeid] = "skipped"
else:
outcomes[nodeid] = "passed"
return {n: (o if o != "missing" or n not in present else "failed") for n, o in outcomes.items()}
def green_types(types: Mapping[str, Any], outcomes: Mapping[str, str]) -> dict[int, str]:
"""``type -> ""`` when green, else the reason. Partial is no: every registered test passes."""
result: dict[int, str] = {}
for key, spec in types.items():
evidence = list(spec.get("evidence", ()))
if not evidence:
result[int(key)] = "ingen probe registrert"
continue
bad = [
f"{n.split('::')[-1]}={outcomes.get(n, 'missing')}"
for n in evidence
if outcomes.get(n) != "passed"
]
result[int(key)] = "; ".join(bad)
return result
def score_types(types: Mapping[str, Any], outcomes: Mapping[str, str]) -> Row:
verdicts = green_types(types, outcomes)
k = sum(1 for why in verdicts.values() if not why)
n = len(types)
exceptions = tuple(
f"type {t} ({types[str(t)]['label']}): {why}" for t, why in sorted(verdicts.items()) if why
)
green = ", ".join(str(t) for t, why in sorted(verdicts.items()) if not why) or "ingen"
status = GREEN if k == n else RED
return Row(
"types",
"3 tilbakemeldingstyper med vei inn OG handling",
k,
n,
status,
f"grønne: {green}",
exceptions=exceptions,
)
# ---------------------------------------------------------------------------------------------
# Row 4 — the round 3 report kept
# ---------------------------------------------------------------------------------------------
def kept_ratio(report: Path, kept: Path) -> tuple[int, int, str]:
if not report.is_file() or not kept.is_file():
return 0, 0, "ingen rapport"
lines = [x for x in report.read_text(encoding="utf-8").splitlines() if x.strip()]
if not lines:
return 0, 0, "tom rapport"
pool = Counter(x for x in kept.read_text(encoding="utf-8").splitlines() if x.strip())
same = 0
for line in lines:
if pool[line] > 0:
pool[line] -= 1
same += 1
return same, len(lines), ""
def score_kept(rounds_dir: Path, threshold: float) -> Row:
same, total, why = kept_ratio(
rounds_dir / "3" / "report.md", rounds_dir / "3" / "report.kept.md"
)
title = f"4 runde 3-rapport beholdt (≥ {threshold:.0%} linjer)"
if why:
return Row("kept", title, None, total or None, RED, why)
ok = same >= threshold * total
return Row(
"kept",
title,
same,
total,
GREEN if ok else RED,
f"{same / total:.1%} av ikke-tomme linjer uendret",
)
# ---------------------------------------------------------------------------------------------
# Row 5 — MAF points, each pointing at a type
# ---------------------------------------------------------------------------------------------
def _imports(tree: ast.AST, construct: str, package: str) -> set[str]:
names: set[str] = set()
for node in ast.walk(tree):
if isinstance(node, ast.ImportFrom) and node.module and node.module.startswith(package):
names |= {a.asname or a.name for a in node.names if a.name == construct}
return names
def _referenced(nodes: Iterable[ast.AST], names: set[str]) -> bool:
return any(isinstance(x, ast.Name) and x.id in names for node in nodes for x in ast.walk(node))
def maf_presence(point: Mapping[str, Any], src: Path) -> tuple[bool, bool]:
"""(construct imported from MAF and used somewhere in ``src``, named call site uses it)."""
present = False
for path in sorted(src.glob("*.py")):
tree = ast.parse(path.read_text(encoding="utf-8"))
names = _imports(tree, point["construct"], point["package"])
body = [
n for n in getattr(tree, "body", []) if not isinstance(n, (ast.Import, ast.ImportFrom))
]
if names and _referenced(body, names):
present = True
break
site = point.get("callsite")
if not site:
return present, False
path = src / site["module"]
if not path.is_file():
return present, False
tree = ast.parse(path.read_text(encoding="utf-8"))
names = _imports(tree, point["construct"], point["package"])
scopes = [
n
for n in getattr(tree, "body", [])
if isinstance(n, (ast.FunctionDef, ast.AsyncFunctionDef, ast.ClassDef))
and n.name == site["scope"]
]
return present, bool(names) and _referenced(scopes, names)
def score_maf(maf: Mapping[str, Any], type_verdicts: Mapping[int, str], src: Path) -> Row:
points = list(maf["points"])
exceptions: list[str] = []
presence = callsites = pointers = 0
for point in points:
present, site = maf_presence(point, src)
types_green = all(not type_verdicts.get(int(t), "ukjent") for t in point["types"])
presence += present
callsites += site
counts = present and site and types_green
pointers += counts
if not counts:
missing = [
w
for w, ok in (
("presence", present),
("kallsted", site),
(f"type {point['types']} grønn", types_green),
)
if not ok
]
exceptions.append(f"{point['u_id']} {point['construct']}: mangler {', '.join(missing)}")
n = len(points)
diagnostics = (
f"presence {presence} av {n}",
f"kallsted verifisert {callsites} av {n}",
f"med grønn typepeker {pointers} av {n}",
)
title = "5 MAF-punkter med typepeker"
if maf.get("approved") is not True:
return Row(
"maf",
title,
0,
n,
RED,
"M ikke godkjent av operatøren",
exceptions=tuple(exceptions),
diagnostics=diagnostics,
)
return Row(
"maf",
title,
pointers,
n,
GREEN if pointers == n else RED,
f"presence {presence} av {n}",
exceptions=tuple(exceptions),
diagnostics=diagnostics,
)
# ---------------------------------------------------------------------------------------------
# Rows 6 and 7 — the stress round's artefacts
# ---------------------------------------------------------------------------------------------
@dataclass(frozen=True)
class StressMeasure:
validated: int = 0
undeclared: int = 0
undeclared_anywhere: int = 0
named: int = 0
rows: int = 0
commissioned: int = 0
where: str = ""
missing: str = ""
undeclared_ids: tuple[str, ...] = field(default=())
def measure_stress(
evidence: Mapping[str, Any], repo_root: Path, stress_root: Path, bundle_root: Path
) -> StressMeasure:
"""Re-judge every listed outbox with the current judge. Any run that cannot be judged makes the
whole measurement absent — a partial one would carry the wrong denominator."""
from portfolio_optimiser import stress
from portfolio_optimiser.mandate import load_mandate
verdicts = []
contexts: set[str] = set()
for run_spec in evidence["runs"]:
context = repo_root / run_spec["context"]
outbox = stress_root / run_spec["outbox"]
if not outbox.is_dir():
return StressMeasure(where=str(stress_root), missing=f"{outbox} finnes ikke")
declared = stress.read_bundle_declarations(context / "bundle.txt")
wanted = run_spec.get("bundle")
chosen = [d for d in declared if wanted in (None, d["name"], d["bundle_id"])]
if len(chosen) != 1:
return StressMeasure(where=str(stress_root), missing=f"{context}: base ikke entydig")
base = bundle_root / chosen[0]["name"]
try:
verdicts.append(
stress.score_context_set(
context,
outbox,
run_spec["run_id"],
base,
bundle_id=chosen[0]["bundle_id"] if len(declared) > 1 else None,
)
)
except (stress.EmptyMeasurement, OSError, ValueError) as exc:
return StressMeasure(where=str(stress_root), missing=f"{run_spec['run_id']}: {exc}")
contexts.add(run_spec["context"])
approaches = [a for v in verdicts for a in v.approaches]
validated = [a for a in approaches if a.status == "validated"]
undeclared = [a for a in validated if a.requirement_source != "approach"]
commissioned = sum(
len(load_mandate(repo_root / c / "mandate.json").approaches) for c in contexts
)
return StressMeasure(
validated=len(validated),
undeclared=len(undeclared),
undeclared_anywhere=sum(1 for a in validated if a.requirement_source == "absent"),
named=sum(1 for a in approaches if a.named),
rows=len(approaches),
commissioned=commissioned,
where=str(stress_root),
undeclared_ids=tuple(a.approach_id for a in undeclared),
)
def score_undeclared(
probes: Sequence[str], outcomes: Mapping[str, str], m: StressMeasure, label: str
) -> Row:
failing = [
f"{n.split('::')[-1]}={outcomes.get(n, 'missing')}"
for n in probes
if outcomes.get(n) != "passed"
]
title = "6 validert UTEN erklært krav (tilnærmingens egen)"
exceptions = [f"probe {x}" for x in failing]
if m.missing:
reason = f"{label}: ikke målt, artefakter mangler ({m.missing})"
k: int | None = None
n: int | None = None
else:
k, n = m.undeclared, m.validated
reason = (
f"{label} ({m.where}): {k} av {n} validerte uten erklæring fra tilnærmingen; "
f"{m.undeclared_anywhere} uten noen erklæring i kjøringen"
)
exceptions += [f"validert uten erklæring: {a}" for a in m.undeclared_ids]
red = bool(failing) or bool(k)
if not probes:
red, exceptions = True, [*exceptions, "ingen probe registrert"]
return Row(
"undeclared", title, k, n, RED if red else GREEN, reason, exceptions=tuple(exceptions)
)
NAMED_WARNING = (
"en prompt-endring som ber modellen gjengi `ref` gjør `named` til noe modellen blir BEDT om, "
"og svekker den som uavhengig måling"
)
def score_named(m: StressMeasure, label: str) -> Row:
title = "7 named (diagnose, ingen terskel)"
if m.missing:
return Row(
"named",
title,
None,
None,
DIAGNOSIS,
f"{label}: ikke målt, artefakter mangler ({m.missing})",
failing=False,
diagnostics=(NAMED_WARNING,),
)
extra = "" if m.rows == m.commissioned else f"; {m.rows} rader dømt"
return Row(
"named",
title,
m.named,
m.commissioned,
DIAGNOSIS,
f"{label}: nevner = bestilte tilnærminger i fasitsettene{extra}",
failing=False,
diagnostics=(NAMED_WARNING,),
)
# ---------------------------------------------------------------------------------------------
# The command
# ---------------------------------------------------------------------------------------------
def evaluate(
*,
rounds_dir: Path,
config: Mapping[str, Any],
repo_root: Path = _REPO_ROOT,
src: Path = _PACKAGE_SRC,
stress_root: Path | None = None,
bundle_root: Path | None = None,
probe_runner: ProbeRunner | None = None,
stress_measure: StressMeasure | None = None,
) -> list[Row]:
ai = ai_authored_lines(repo_root, config["ai_authored"])
required = int(config["rounds_required"])
types = config["feedback_types"]
probes = list(config["row6_evidence"])
nodeids = [n for spec in types.values() for n in spec.get("evidence", ())] + probes
outcomes = (probe_runner or (lambda ids: run_probes(ids, repo_root)))(nodeids)
evidence = config["stress_evidence"]
if stress_measure is None:
stress_measure = measure_stress(
evidence,
repo_root,
stress_root or repo_root / evidence["root"],
bundle_root
or Path(
os.environ.get("PORTFOLIO_VEGNORMAL_ROOT", "~/repos/vegnormal-okf/build/ferdig")
).expanduser(),
)
return [
score_rounds(rounds_dir, required, ai),
score_changes(rounds_dir, required, ai),
score_types(types, outcomes),
score_kept(rounds_dir, float(config["keep_threshold"])),
score_maf(config["maf_points"], green_types(types, outcomes), src),
score_undeclared(probes, outcomes, stress_measure, evidence["label"]),
score_named(stress_measure, evidence["label"]),
]
def exit_code(rows: Sequence[Row]) -> int:
return 0 if all(r.status == GREEN for r in rows if r.failing) else 1
def render(rows: Sequence[Row]) -> str:
out = ["rad | k av N | status | grunn"]
out += [r.line() for r in rows]
out.append("")
out.append("Unntak fra 100 %:")
for r in rows:
for x in r.exceptions:
out.append(f" [{r.title.split()[0]}] {x}")
for d in r.diagnostics:
out.append(f" [{r.title.split()[0]}] diagnose: {d}")
return "\n".join(out)
def main(argv: Sequence[str] | None = None) -> int:
parser = argparse.ArgumentParser(
prog="python -m portfolio_optimiser.evals.v1_gate",
description="Hvor langt portfolio-optimiser er fra v1, rad for rad. Exit 0 kun når alle "
"fellende rader (1-6) er grønne, 1 ellers, 2 ved feil bruk. Ingen modellkall, intet nett.",
epilog=ROUNDS_CONTRACT,
formatter_class=argparse.RawDescriptionHelpFormatter,
)
parser.add_argument(
"--rounds-dir",
default=None,
help=f"rundekatalogen (default {DEFAULT_ROUNDS_DIR}/, gitignored)",
)
parser.add_argument(
"--stress-root", default=None, help="utboks-roten for stressrunden rad 6-7 dømmer"
)
parser.add_argument(
"--bundle-root", default=None, help="der stressrundens kunnskapsbaser er montert"
)
parser.add_argument("--json", action="store_true", help="maskinlesbar output")
args = parser.parse_args(argv)
if args.rounds_dir is not None and not Path(args.rounds_dir).is_dir():
parser.error(f"--rounds-dir {args.rounds_dir} finnes ikke")
rounds_dir = Path(args.rounds_dir) if args.rounds_dir else _REPO_ROOT / DEFAULT_ROUNDS_DIR
rows = evaluate(
rounds_dir=rounds_dir,
config=load_config(),
stress_root=Path(args.stress_root) if args.stress_root else None,
bundle_root=Path(args.bundle_root).expanduser() if args.bundle_root else None,
)
code = exit_code(rows)
if args.json:
print(
json.dumps(
{"exit": code, "rows": [asdict(r) for r in rows]}, ensure_ascii=False, indent=2
)
)
else:
print(render(rows))
return code
if __name__ == "__main__": # pragma: no cover - exercised by a subprocess test
raise SystemExit(main())