feat: 回忆录证据血缘与内部评测可追溯,顺带对齐本地评测台与 CI

数据库与模型:新增多版迁移(章节证据快照、对话血缘、记忆事实/时间线 lineage 等),把「成稿 ↔ 对话/记忆」的溯源信息落到表结构里。
业务链路:会话与 WS、回忆录/故事流水线、记忆写入与 enrichment 等跟着接上线索与快照;新增章节证据快照与评测侧 EvalTraceService 等模块,方便组评审用的证据包。
内部评测:自动化 run 与手工 memoir 评审共用可追溯证据;rubric/ judge 相关脚本与文档有配套调整。
app-eval-web:Memoir/实验详情里能展开看证据摘要与 evidence_trace(含对话轮次 id);Vite 代理与 development.sh 注入的 API 端口与当前默认内部评测端口一致,避免改端口后页面连错服务。
工程杂项:GitHub Actions / 仓库说明有更新;各适配器与支付/配额/plan 等多处为小改动或跟随主改动的收尾;新增/扩充了?
This commit is contained in:
Kevin
2026-04-08 15:37:09 +08:00
parent 6772e1269c
commit 309a051038
109 changed files with 4125 additions and 858 deletions

View File

@@ -13,6 +13,8 @@ import re
from dataclasses import dataclass
from typing import Any
from pydantic import ValidationError
from app.agents.memoir.prompts import get_chapter_classification_json_prompt
from app.agents.memoir.schemas import ClassificationOutput
from app.agents.stage_constants import (
@@ -22,7 +24,7 @@ from app.agents.stage_constants import (
)
from app.core.config import settings
from app.core.json_utils import extract_json_payload
from app.core.llm_call import llm_json_call
from app.core.llm_call import LLMCallError, llm_json_call
from app.core.logging import get_logger
logger = get_logger(__name__)
@@ -159,7 +161,7 @@ class ClassificationAgent:
)
if category in CHAPTER_CATEGORIES:
return ChapterClassifyResult(category=category, llm_said_none=False)
except Exception as e:
except (LLMCallError, ValidationError, ValueError, KeyError) as e:
logger.warning("ClassificationAgent LLM 章节分类失败: {}", e)
stage = _detect_stage(text, fallback_stage)