相对第一版 46fc976 的完整变更。组员迁移对照表见 docs/20。
一、对外契约对齐 docs/05(破坏性,共 4 处,组员需按 docs/20 调整)
1) 配置发布端点改为文档规定的复数资源名:submit→validations、
approve→reviews(需 body decision)、activate→activations、
rollback→rollbacks;第一版这 4 个动词式路径 docs/05 从未定义过。
2) 错误码由 8 个笼统码改为 15 个具体语义码(FORBIDDEN→AGENT_PERMISSION_DENIED、
UNAUTHORIZED→AUTHENTICATION_REQUIRED、CONFLICT→RESOURCE_VERSION_CONFLICT、
RESOURCE_NOT_FOUND→RUN_NOT_FOUND/SESSION_NOT_FOUND 等),
输入类错误状态码 400→422。
3) POST /api/v1/agent-runs 与 GET /api/v1/agent-runs/{run_id} 统一为
{data, meta} 信封(data 内字段名与语义未变)。
4) 错误响应体统一为 {error:{code,message,retryable,field_errors}, meta:{trace_id}},
不再返回 FastAPI 默认的 {"detail": ...}。
二、数据库基线与约束
新增 39 张表的基线迁移(链根)与联合唯一键纠偏(4 张表、删 8 增 4,幂等收敛);
撤下 config_release 的双人复核 CHECK(应用层已允许自审,审核节点保留,
自审如实写入 reviewer_id);记忆 active key 生成列与唯一键;
activate 开始记录 supersedes_release_id 使版本链可追溯。
docs/00 基线未修改,未重命名或删除任何表与字段。
三、修复会静默出错或无报错的缺陷
- 跑完集成测试后平台会静默失去生效配置:清理只删自己创建的版本,却没有恢复被它
顶成 superseded 的原生效版本,且审计一并删除因而完全无痕,表现为所有工具被拒
但没有任何报错。已修清理逻辑并加恢复。
- Worker 单轮异常导致进程退出;记忆抽取调用方的“事务已开始”异常;
召回缓存丢失 degraded 标记;连接时区未生效导致 created_at/updated_at 差 8 小时;
.env 与 os.getenv 密钥来源分裂导致“没有可用的已批准模型端点”。
- 记忆信号识别漏判与跨键误命中;SSE 未带 Accept 的协商行为。
四、功能补齐
记忆链路 P1/P2/P3(抽取、受控词表、召回与缓存、生命周期级联及投影事件)、
fin_* 场内交易只读 ORM 层、agent_intent_config 状态流转并在运行期真正生效、
限流(Redis 固定窗口、故障一律放行)、游标校验、trace_id 中间件、
示例业务 Agent fund_query_demo 与一键端到端验证脚本,以及审计/指纹/迁移状态工具。
五、文档与验证
新增 docs/19(业务 Agent 接入实操)、docs/20(第一版迁移指南)与 docs/evidence 证据;
docs/01/02/06/08/09/17 同步实现现状。
验证结果:ruff 通过、mypy 103 文件无错、unit+contract 447 passed、
integration 29 passed、acceptance_check --production 7 PASS、
demo_agent_e2e 9/9 PASS(含失败关闭反证)。
108 lines
5.6 KiB
Python
108 lines
5.6 KiB
Python
import asyncio
|
|
from uuid import uuid4
|
|
|
|
import httpx
|
|
import pytest
|
|
from sqlalchemy import delete, select, update
|
|
|
|
from app.api.dependencies.auth import build_request_context
|
|
from app.core.contracts import AgentRequest, AgentResult, CoreResult, RequestContext
|
|
from app.core.errors import RunLeaseLostError
|
|
from app.infrastructure.db import SessionFactory
|
|
from app.main import create_app
|
|
from app.model.audit import InteractionAudit
|
|
from app.model.conversation import ConversationMessage
|
|
from app.model.platform import AgentRun, DomainEventOutbox, OutboxDelivery, RequestIdempotency
|
|
from app.service.agent_persistence_service import AgentPersistenceService
|
|
from app.service.agent_run_application_service import AgentRunApplicationService
|
|
from app.worker.runtime import WorkerRuntime
|
|
|
|
|
|
@pytest.mark.integration
|
|
@pytest.mark.parametrize("revoked", [False, True])
|
|
async def test_http_accept_worker_commit_query_and_repeat(acceptance_registry, revoked):
|
|
context = RequestContext(user_id="1", trace_id=str(uuid4()), roles=("customer",),
|
|
permissions=("agent:run",))
|
|
app = create_app()
|
|
app.dependency_overrides[build_request_context] = lambda: context
|
|
session_id = f"runtime-{uuid4()}"
|
|
run_id = None
|
|
|
|
async def identity(value):
|
|
return context if not revoked else value
|
|
|
|
runtime = WorkerRuntime(acceptance_registry, resolve_identity=identity)
|
|
async with httpx.AsyncClient(transport=httpx.ASGITransport(app=app),
|
|
base_url="http://test") as client:
|
|
try:
|
|
# 消息必须命中显式记忆信号:P2 之后触发判定改为事件/事实驱动,
|
|
# 无信号的普通消息不再产生 memory.extraction_requested,而下方断言依赖该事件。
|
|
response = await client.post("/api/v1/agent-runs", json={
|
|
"agent_type": "customer_service",
|
|
"message": "hello runtime,我的风险偏好是稳健型",
|
|
"session_id": session_id, "idempotency_key": str(uuid4()),
|
|
})
|
|
assert response.status_code == 202
|
|
run_id = response.json()["data"]["run_id"]
|
|
assert await runtime.dispatch_one(run_id=run_id)
|
|
assert not await runtime.dispatch_one(run_id=run_id)
|
|
await asyncio.gather(runtime.execute(run_id), runtime.execute(run_id))
|
|
result = (await client.get(f"/api/v1/agent-runs/{run_id}")).json()["data"]
|
|
assert result["status"] == ("failed" if revoked else "succeeded")
|
|
if not revoked:
|
|
assert result["result"]["content"] == "test result"
|
|
assert not await runtime.execute(run_id)
|
|
async with SessionFactory() as session:
|
|
messages = list(await session.scalars(select(ConversationMessage).where(
|
|
ConversationMessage.session_id == session_id,
|
|
ConversationMessage.role == "assistant")))
|
|
assert len(messages) == (0 if revoked else 1)
|
|
events = list(await session.scalars(select(DomainEventOutbox).where(
|
|
DomainEventOutbox.aggregate_id == run_id,
|
|
DomainEventOutbox.event_type == "memory.extraction_requested")))
|
|
assert len(events) == (0 if revoked else 1)
|
|
finally:
|
|
await cleanup(session_id, run_id)
|
|
|
|
|
|
async def test_expired_worker_cannot_commit(acceptance_registry):
|
|
context = RequestContext(user_id="1", trace_id=str(uuid4()), roles=("customer",),
|
|
permissions=("agent:run",))
|
|
request = AgentRequest(agent_type="customer_service", message="fence test",
|
|
session_id=f"fence-{uuid4()}", idempotency_key=str(uuid4()))
|
|
async with SessionFactory() as session:
|
|
accepted = await AgentRunApplicationService(session, acceptance_registry).accept(
|
|
request, context)
|
|
try:
|
|
async with SessionFactory() as session, session.begin():
|
|
await session.execute(update(AgentRun).where(AgentRun.run_id == accepted.run_id)
|
|
.values(status="running", worker_id="new-owner"))
|
|
async with SessionFactory() as session:
|
|
with pytest.raises(RunLeaseLostError):
|
|
await AgentPersistenceService(session).complete_run(
|
|
accepted.run_id, AgentResult(run_id=accepted.run_id,
|
|
result=CoreResult(text="stale")),
|
|
worker_id="old-owner")
|
|
async with SessionFactory() as session:
|
|
assert await session.scalar(select(ConversationMessage.id).where(
|
|
ConversationMessage.session_id == request.session_id,
|
|
ConversationMessage.role == "assistant")) is None
|
|
finally:
|
|
await cleanup(request.session_id, accepted.run_id)
|
|
|
|
|
|
async def cleanup(session_id, run_id):
|
|
async with SessionFactory() as session, session.begin():
|
|
event_ids = select(DomainEventOutbox.event_id).where(
|
|
DomainEventOutbox.aggregate_id == run_id)
|
|
await session.execute(delete(OutboxDelivery).where(OutboxDelivery.event_id.in_(event_ids)))
|
|
await session.execute(delete(DomainEventOutbox).where(
|
|
DomainEventOutbox.aggregate_id == run_id))
|
|
await session.execute(delete(InteractionAudit).where(
|
|
InteractionAudit.session_id == session_id))
|
|
await session.execute(delete(AgentRun).where(AgentRun.session_id == session_id))
|
|
await session.execute(delete(RequestIdempotency).where(
|
|
RequestIdempotency.session_id == session_id))
|
|
await session.execute(delete(ConversationMessage).where(
|
|
ConversationMessage.session_id == session_id))
|