G0~G8 성과·동맹 측정 OS 작업 일괄 고정
8월 7일까지 워킹트리에만 남아 있던 미커밋 작업을 커밋한다. 여러 사본 폴더(worktree·clone)에 흩어져 있던 중간 스냅샷을 정리하기 전에 원본을 git 이력으로 고정하는 것이 목적이다. - contracts/routes/services: measurement, outcome_trajectory, rupture_repair, deliberate_practice, calibration_transfer, supervision_research, multimodal_alliance, continuous_improvement 계열 신규 모듈과 테스트 - infra/db/init: 07~16 마이그레이션(측정 기반~calibration transfer 실행) - apps/web: 세션 리뷰 카드·관리 화면·E2E 스펙 추가 - docs/ops: G0~G8 라이브 통합·배포·롤백 증거 문서와 evidence JSON/PNG - scripts: smoke·ledger·릴리스 에이전트·NAS 프리뷰 운영 스크립트 engine.public 로그 .bak과 apps/web/test-results 산출물은 커밋에서 제외했다.
This commit is contained in:
Yun Chan
parent
93dd8f82d7
commit
16e791e044
390 changed files with 243188 additions and 499 deletions
442
scripts/smoke-outcome-trajectory-api.py
Normal file
442
scripts/smoke-outcome-trajectory-api.py
Normal file
|
|
@ -0,0 +1,442 @@
|
|||
"""Exercise the complete G2 five-session outcome-trajectory HTTP flow.
|
||||
|
||||
The smoke intentionally retains its unique dev/E2E fixture. It proves case
|
||||
continuity, three independent outcome axes, retry idempotency, cross-role and
|
||||
cross-cohort access boundaries, role-safe relationship memory, and the
|
||||
non-clinical synthetic-arc notice without printing credentials, cookies,
|
||||
e-mail addresses, or transcript text.
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import json
|
||||
import secrets
|
||||
import time
|
||||
import urllib.error
|
||||
import urllib.request
|
||||
from dataclasses import dataclass
|
||||
from http.cookiejar import CookieJar
|
||||
from pathlib import Path
|
||||
from typing import Any
|
||||
from uuid import uuid4
|
||||
|
||||
|
||||
AXES = ("distress_load", "daily_functioning", "learning_engagement")
|
||||
EXPECTED_SCORES = (
|
||||
{"distress_load": 0.70, "daily_functioning": 0.30, "learning_engagement": 0.40},
|
||||
{"distress_load": 0.62, "daily_functioning": 0.40, "learning_engagement": 0.48},
|
||||
{"distress_load": 0.54, "daily_functioning": 0.50, "learning_engagement": 0.56},
|
||||
{"distress_load": 0.46, "daily_functioning": 0.60, "learning_engagement": 0.64},
|
||||
{"distress_load": 0.38, "daily_functioning": 0.68, "learning_engagement": 0.72},
|
||||
)
|
||||
COHORT_ID = "e2e-hanshin"
|
||||
|
||||
|
||||
class SmokeError(RuntimeError):
|
||||
pass
|
||||
|
||||
|
||||
@dataclass(frozen=True)
|
||||
class ApiResponse:
|
||||
status: int
|
||||
body: Any
|
||||
|
||||
|
||||
class ApiClient:
|
||||
def __init__(self, base_url: str, timeout: float) -> None:
|
||||
self.base_url = base_url.rstrip("/")
|
||||
self.timeout = timeout
|
||||
self._opener = urllib.request.build_opener(
|
||||
urllib.request.HTTPCookieProcessor(CookieJar())
|
||||
)
|
||||
|
||||
def request(
|
||||
self,
|
||||
method: str,
|
||||
path: str,
|
||||
payload: dict[str, Any] | None = None,
|
||||
*,
|
||||
expected: set[int] | None = None,
|
||||
) -> ApiResponse:
|
||||
data = None
|
||||
headers = {"Accept": "application/json"}
|
||||
if payload is not None:
|
||||
data = json.dumps(payload, ensure_ascii=False).encode("utf-8")
|
||||
headers["Content-Type"] = "application/json"
|
||||
request = urllib.request.Request(
|
||||
f"{self.base_url}{path}", data=data, headers=headers, method=method
|
||||
)
|
||||
try:
|
||||
with self._opener.open(request, timeout=self.timeout) as response:
|
||||
raw = response.read().decode("utf-8")
|
||||
result = ApiResponse(
|
||||
status=response.status,
|
||||
body=json.loads(raw) if raw else {},
|
||||
)
|
||||
except urllib.error.HTTPError as exc:
|
||||
raw = exc.read().decode("utf-8", errors="replace")
|
||||
try:
|
||||
body = json.loads(raw) if raw else {}
|
||||
except json.JSONDecodeError:
|
||||
body = {"detail": raw[:500]}
|
||||
result = ApiResponse(status=exc.code, body=body)
|
||||
except urllib.error.URLError as exc:
|
||||
raise SmokeError(
|
||||
f"{method} {path} transport failed: {type(exc.reason).__name__}"
|
||||
) from exc
|
||||
|
||||
allowed = expected or {200}
|
||||
if result.status not in allowed:
|
||||
detail = (
|
||||
result.body.get("detail") if isinstance(result.body, dict) else None
|
||||
)
|
||||
raise SmokeError(
|
||||
f"{method} {path} returned HTTP {result.status}; detail={detail!r}"
|
||||
)
|
||||
return result
|
||||
|
||||
|
||||
def _onboarding_payload(display_name: str) -> dict[str, Any]:
|
||||
return {
|
||||
"legal_name": display_name,
|
||||
"affiliation": "한신대학교",
|
||||
"department": "상담심리학과",
|
||||
"grade_level": "통합검증",
|
||||
"phone": "010-0000-0000",
|
||||
"contact_address": "경기도 오산시 한신대학교",
|
||||
"nickname": display_name,
|
||||
"self_introduction": "G2 5회기 성과 궤적 API를 검증하는 개발 fixture입니다.",
|
||||
"avatar_url": "",
|
||||
"terms_accepted": True,
|
||||
"privacy_accepted": True,
|
||||
}
|
||||
|
||||
|
||||
def _sign_in(
|
||||
client: ApiClient,
|
||||
*,
|
||||
suffix: str,
|
||||
identity: str,
|
||||
role: str,
|
||||
cohort_ids: list[str],
|
||||
) -> None:
|
||||
login = client.request(
|
||||
"POST",
|
||||
"/auth/dev-login",
|
||||
{
|
||||
"email": f"dev.e2e.outcome.{identity}.{suffix}@hs.ac.kr",
|
||||
"role": role,
|
||||
"display_name": f"Outcome {identity.title()}",
|
||||
"cohort_ids": cohort_ids,
|
||||
},
|
||||
)
|
||||
if login.body.get("role") != role:
|
||||
raise SmokeError(f"dev-login role mismatch for {identity}")
|
||||
client.request(
|
||||
"POST",
|
||||
"/users/me/onboarding",
|
||||
_onboarding_payload(f"Outcome {identity.title()}"),
|
||||
)
|
||||
|
||||
|
||||
def _choose_persona(client: ApiClient) -> str:
|
||||
response = client.request("GET", "/personas")
|
||||
if not isinstance(response.body, list):
|
||||
raise SmokeError("persona catalog did not return a list")
|
||||
usable = [
|
||||
item
|
||||
for item in response.body
|
||||
if isinstance(item, dict)
|
||||
and item.get("source") == "database"
|
||||
and not item.get("degraded")
|
||||
and item.get("code")
|
||||
]
|
||||
if not usable:
|
||||
raise SmokeError("persona catalog has no non-degraded database persona")
|
||||
preferred = next((item for item in usable if item.get("code") == "P1"), usable[0])
|
||||
return str(preferred["code"])
|
||||
|
||||
|
||||
def _assert_no_total_score(value: Any, path: str = "response") -> None:
|
||||
if isinstance(value, dict):
|
||||
forbidden = {"total", "total_score", "overall_score"} & set(value)
|
||||
if forbidden:
|
||||
raise SmokeError(f"{path} exposed forbidden total score keys: {forbidden}")
|
||||
for key, child in value.items():
|
||||
_assert_no_total_score(child, f"{path}.{key}")
|
||||
elif isinstance(value, list):
|
||||
for index, child in enumerate(value):
|
||||
_assert_no_total_score(child, f"{path}[{index}]")
|
||||
|
||||
|
||||
def _assert_trajectory(payload: dict[str, Any], *, session_count: int) -> None:
|
||||
expected_arc = payload.get("expected_arc") or {}
|
||||
assessment = payload.get("assessment") or {}
|
||||
sessions = assessment.get("sessions") or []
|
||||
observations = payload.get("observations") or []
|
||||
if expected_arc.get("clinical_claim_allowed") is not False:
|
||||
raise SmokeError("synthetic expected arc did not prohibit clinical claims")
|
||||
if expected_arc.get("data_classification") != "synthetic_educational":
|
||||
raise SmokeError("expected arc omitted synthetic educational classification")
|
||||
if len(expected_arc.get("distributions") or []) != 15:
|
||||
raise SmokeError("expected arc is not 5 sessions x 3 independent axes")
|
||||
if assessment.get("clinical_claim_allowed") is not False:
|
||||
raise SmokeError("assessment did not prohibit clinical claims")
|
||||
if len(sessions) != session_count:
|
||||
raise SmokeError(
|
||||
f"trajectory contains {len(sessions)} sessions, expected {session_count}"
|
||||
)
|
||||
if len(observations) != session_count * 3:
|
||||
raise SmokeError(
|
||||
f"trajectory contains {len(observations)} observations, "
|
||||
f"expected {session_count * 3}"
|
||||
)
|
||||
for expected_no, item in enumerate(sessions, start=1):
|
||||
if item.get("session_no") != expected_no or item.get("status") != "on_track":
|
||||
raise SmokeError(f"session trajectory differs at S{expected_no}: {item}")
|
||||
axes = item.get("axes") or []
|
||||
if len(axes) != 3 or {axis.get("axis") for axis in axes} != set(AXES):
|
||||
raise SmokeError(f"S{expected_no} did not preserve three independent axes")
|
||||
_assert_no_total_score(payload)
|
||||
|
||||
|
||||
def _first_durable_turn_id(review: dict[str, Any]) -> str:
|
||||
turns = review.get("turns") or []
|
||||
durable_ids = [str(item.get("turn_id")) for item in turns if item.get("turn_id")]
|
||||
if not durable_ids:
|
||||
raise SmokeError("session review omitted durable turn UUID evidence")
|
||||
return durable_ids[0]
|
||||
|
||||
|
||||
def run(args: argparse.Namespace) -> dict[str, Any]:
|
||||
health_client = ApiClient(args.api_base_url, args.request_timeout)
|
||||
health = health_client.request("GET", "/health")
|
||||
if not health.body.get("db") or not health.body.get("engine"):
|
||||
raise SmokeError("API health is not DB+engine ready")
|
||||
|
||||
suffix = f"{int(time.time())}.{secrets.token_hex(3)}"
|
||||
learner = ApiClient(args.api_base_url, args.request_timeout)
|
||||
teacher = ApiClient(args.api_base_url, args.request_timeout)
|
||||
other_learner = ApiClient(args.api_base_url, args.request_timeout)
|
||||
other_teacher = ApiClient(args.api_base_url, args.request_timeout)
|
||||
_sign_in(
|
||||
learner,
|
||||
suffix=suffix,
|
||||
identity="learner",
|
||||
role="learner",
|
||||
cohort_ids=[COHORT_ID],
|
||||
)
|
||||
_sign_in(
|
||||
teacher,
|
||||
suffix=suffix,
|
||||
identity="teacher",
|
||||
role="teacher",
|
||||
cohort_ids=[COHORT_ID],
|
||||
)
|
||||
_sign_in(
|
||||
other_learner,
|
||||
suffix=suffix,
|
||||
identity="other-learner",
|
||||
role="learner",
|
||||
cohort_ids=[COHORT_ID],
|
||||
)
|
||||
_sign_in(
|
||||
other_teacher,
|
||||
suffix=suffix,
|
||||
identity="other-teacher",
|
||||
role="teacher",
|
||||
cohort_ids=["e2e-other-cohort"],
|
||||
)
|
||||
persona_code = _choose_persona(learner)
|
||||
|
||||
case_id = ""
|
||||
session_ids: list[str] = []
|
||||
submission_ids: list[str] = []
|
||||
measurement_ids: list[list[str]] = []
|
||||
first_turn_id = ""
|
||||
final_payload: dict[str, Any] = {}
|
||||
first_retry_stable = False
|
||||
conflict_rejected = False
|
||||
|
||||
for session_no, scores in enumerate(EXPECTED_SCORES, start=1):
|
||||
started = learner.request(
|
||||
"POST",
|
||||
"/sessions",
|
||||
{
|
||||
"persona_code": persona_code,
|
||||
"theory_mode": "humanistic",
|
||||
"goal_stages": ["라포", "탐색"],
|
||||
},
|
||||
expected={201},
|
||||
)
|
||||
if started.body.get("degraded"):
|
||||
raise SmokeError(f"S{session_no} start was degraded")
|
||||
if started.body.get("session_no") != session_no:
|
||||
raise SmokeError(
|
||||
f"session continuity differs: {started.body.get('session_no')} != {session_no}"
|
||||
)
|
||||
current_case_id = str(started.body.get("case_id") or "")
|
||||
if not current_case_id:
|
||||
raise SmokeError("session start omitted case_id")
|
||||
if case_id and current_case_id != case_id:
|
||||
raise SmokeError("five sessions did not remain in one case")
|
||||
case_id = current_case_id
|
||||
session_id = str(started.body.get("session_id") or "")
|
||||
session_ids.append(session_id)
|
||||
|
||||
if session_no == 1:
|
||||
turn = learner.request(
|
||||
"POST",
|
||||
f"/sessions/{session_id}/turn",
|
||||
{"text": "오늘 확인할 목표를 함께 정해도 괜찮을까요?"},
|
||||
)
|
||||
if not turn.body.get("client_reply"):
|
||||
raise SmokeError("first-session turn omitted client reply")
|
||||
|
||||
learner.request("POST", f"/sessions/{session_id}/end")
|
||||
if session_no == 1:
|
||||
review = learner.request("GET", f"/sessions/{session_id}/review")
|
||||
first_turn_id = _first_durable_turn_id(review.body)
|
||||
|
||||
submission_id = str(uuid4())
|
||||
submission_ids.append(submission_id)
|
||||
submission = {
|
||||
"submission_id": submission_id,
|
||||
"scores": scores,
|
||||
"confidences": {axis: 0.9 for axis in AXES},
|
||||
"evidence_turn_ids": [first_turn_id] if session_no == 1 else [],
|
||||
}
|
||||
created = learner.request(
|
||||
"POST",
|
||||
f"/sessions/{session_id}/outcome-observations",
|
||||
submission,
|
||||
expected={201},
|
||||
)
|
||||
_assert_trajectory(created.body, session_count=session_no)
|
||||
ids = [
|
||||
str(item) for item in created.body.get("submitted_measurement_ids") or []
|
||||
]
|
||||
if len(ids) != 3:
|
||||
raise SmokeError(f"S{session_no} did not return three measurement IDs")
|
||||
measurement_ids.append(ids)
|
||||
final_payload = created.body
|
||||
|
||||
if session_no == 1:
|
||||
retried = learner.request(
|
||||
"POST",
|
||||
f"/sessions/{session_id}/outcome-observations",
|
||||
submission,
|
||||
expected={201},
|
||||
)
|
||||
first_retry_stable = retried.body.get("submitted_measurement_ids") == ids
|
||||
if not first_retry_stable:
|
||||
raise SmokeError("same submission retry changed measurement IDs")
|
||||
changed = dict(submission)
|
||||
changed["scores"] = dict(scores, distress_load=0.99)
|
||||
conflict = learner.request(
|
||||
"POST",
|
||||
f"/sessions/{session_id}/outcome-observations",
|
||||
changed,
|
||||
expected={409},
|
||||
)
|
||||
conflict_rejected = conflict.status == 409
|
||||
|
||||
final_session_id = session_ids[-1]
|
||||
relationship = teacher.request(
|
||||
"POST",
|
||||
f"/sessions/{session_ids[0]}/relationship-memory-events",
|
||||
{
|
||||
"event_type": "goal_agreement",
|
||||
"summaries": {
|
||||
"counselor": "첫 회기의 학습 목표를 내담자와 명시적으로 합의했다.",
|
||||
"supervisor": "목표 합의 발화가 실제 축어록 근거로 확인됐다.",
|
||||
},
|
||||
"evidence_turn_ids": [first_turn_id],
|
||||
},
|
||||
expected={201},
|
||||
)
|
||||
if relationship.body.get("status") != "recorded":
|
||||
raise SmokeError("relationship memory was not recorded")
|
||||
|
||||
teacher_read = teacher.request(
|
||||
"GET", f"/sessions/{final_session_id}/outcome-trajectory"
|
||||
)
|
||||
_assert_trajectory(teacher_read.body, session_count=5)
|
||||
relationship_memory = teacher_read.body.get("relationship_memory") or []
|
||||
if len(relationship_memory) != 1:
|
||||
raise SmokeError("teacher trajectory omitted role-safe relationship memory")
|
||||
if set(relationship_memory[0]) & {"summaries", "client", "evaluator"}:
|
||||
raise SmokeError(
|
||||
"teacher relationship projection leaked another role's summary"
|
||||
)
|
||||
|
||||
learner_read = learner.request(
|
||||
"GET", f"/sessions/{final_session_id}/outcome-trajectory"
|
||||
)
|
||||
_assert_trajectory(learner_read.body, session_count=5)
|
||||
if len(learner_read.body.get("relationship_memory") or []) != 1:
|
||||
raise SmokeError("learner trajectory omitted counselor relationship memory")
|
||||
|
||||
other_learner.request(
|
||||
"GET", f"/sessions/{final_session_id}/outcome-trajectory", expected={404}
|
||||
)
|
||||
other_teacher.request(
|
||||
"GET", f"/sessions/{final_session_id}/outcome-trajectory", expected={404}
|
||||
)
|
||||
|
||||
return {
|
||||
"ok": True,
|
||||
"api_base_url": args.api_base_url,
|
||||
"health": {
|
||||
"status": health.body.get("status"),
|
||||
"db": health.body.get("db"),
|
||||
"engine": health.body.get("engine"),
|
||||
"engine_mode": health.body.get("engine_mode"),
|
||||
},
|
||||
"fixture_policy": "retained unique dev:e2e identities; no fixture deletion",
|
||||
"cohort_id": COHORT_ID,
|
||||
"persona_code": persona_code,
|
||||
"case_id": case_id,
|
||||
"session_ids": session_ids,
|
||||
"submission_ids": submission_ids,
|
||||
"proof": {
|
||||
"session_numbers": [1, 2, 3, 4, 5],
|
||||
"same_case_across_sessions": True,
|
||||
"durable_review_turn_uuid_exposed": bool(first_turn_id),
|
||||
"observed_axis_count": len(final_payload.get("observations") or []),
|
||||
"session_statuses": [
|
||||
item.get("status")
|
||||
for item in (final_payload.get("assessment") or {}).get("sessions", [])
|
||||
],
|
||||
"three_axes_without_total_score": True,
|
||||
"synthetic_non_clinical_notice": True,
|
||||
"same_submission_measurement_ids_stable": first_retry_stable,
|
||||
"changed_payload_same_submission_rejected": conflict_rejected,
|
||||
"measurement_id_count": sum(len(items) for items in measurement_ids),
|
||||
"teacher_cohort_read": True,
|
||||
"other_learner_rejected": True,
|
||||
"cross_cohort_teacher_rejected": True,
|
||||
"role_safe_relationship_memory_count": len(relationship_memory),
|
||||
},
|
||||
}
|
||||
|
||||
|
||||
def main() -> None:
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("--api-base-url", default="http://127.0.0.1:8005")
|
||||
parser.add_argument("--request-timeout", type=float, default=180.0)
|
||||
parser.add_argument("--out", default="")
|
||||
args = parser.parse_args()
|
||||
|
||||
result = run(args)
|
||||
text = json.dumps(result, ensure_ascii=False, indent=2)
|
||||
if args.out:
|
||||
out_path = Path(args.out)
|
||||
out_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
out_path.write_text(text + "\n", encoding="utf-8")
|
||||
print(text)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
Loading…
Add table
Add a link
Reference in a new issue