ab1589921a
## 安全修复 (12项) - Webhook接口添加全局Token认证，过滤敏感请求头 - 修复JWT Base64 padding公式，防止签名验证绕过 - 数据库密码/飞书Token从源码移除，改为环境变量 - 工作流引擎添加路径遍历防护 (_resolve_safe_path) - eval()添加模板长度上限检查 - 审批API添加认证依赖 - 前端v-html增强XSS转义，console.log仅开发模式输出 - 500错误不再暴露内部异常详情 ## Agent运行时修复 (7项) - 删除_inject_knowledge_context中未定义db变量的finally块 - 工具执行添加try/except保护，异常不崩溃Agent - LLM重试计入budget计数器 - self_review异常时passed=False - max_iterations截断标记success=False - 工具参数JSON解析失败时记录警告日志 - run()开始时重置_llm_invocations计数器 ## 配置与基础设施 - DEBUG默认False，SQL_ECHO独立配置项 - init_db()补全13个缺失模型导入 - 新增WEBHOOK_AUTH_TOKEN/SQL_ECHO配置项 - 新增.env.example模板文件 ## 前端修复 (12项) - 登录改用URLSearchParams替代FormData - 401拦截器通过Pinia store统一清理状态 - SSE流超时从60s延长至300s - final/error事件时清除streamTimeout - localStorage聊天记录添加24h TTL - safeParseArgCount替代模板中裸JSON.parse - fetchUser 401时同时清除user对象 ## 新增模块 - 知识进化: knowledge_extractor/retriever/tasks - 数字孪生: shadow_executor/comparison模型 - 行为采集: behavior_middleware/collector/fingerprint_engine - 代码审查: code_review_agent/document_review_agent - 反馈学习: feedback_learner - 瓶颈检测/优化引擎/成本估算/需求估算 - 速率限制器 (rate_limiter) - Alembic迁移 015-020 ## 文档 - 商业化落地计划 - 8篇docs文档 (架构/API/部署/开发/贡献等) - Docker Compose生产配置 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
68 lines
2.9 KiB
Python
68 lines
2.9 KiB
Python
"""
|
|
知识条目模型 — Agent 执行经验的结构化沉淀
|
|
"""
|
|
import uuid
|
|
from datetime import datetime
|
|
from sqlalchemy import Column, String, Text, Integer, DateTime, Boolean, JSON, Float, Index
|
|
from app.core.database import Base
|
|
|
|
|
|
class KnowledgeEntry(Base):
|
|
"""从 Agent 执行日志中提取的可复用知识条目"""
|
|
__tablename__ = "knowledge_entries"
|
|
|
|
id = Column(String(36), primary_key=True, default=lambda: str(uuid.uuid4()))
|
|
title = Column(String(500), nullable=False, comment="知识标题（一句话概括）")
|
|
category = Column(String(30), nullable=False, index=True,
|
|
comment="类别: bug_fix/best_practice/workaround/optimization/insight")
|
|
tags = Column(JSON, nullable=True, comment="标签列表: ['mysql','deadlock','retry']")
|
|
|
|
# 知识内容
|
|
situation = Column(Text, nullable=True, comment="适用场景")
|
|
solution = Column(Text, nullable=True, comment="解决方案")
|
|
caveats = Column(Text, nullable=True, comment="注意事项/踩坑记录")
|
|
|
|
# 来源追溯
|
|
source_execution_ids = Column(JSON, nullable=True, comment="原始执行日志ID列表")
|
|
source_agent_name = Column(String(200), nullable=True, comment="来源 Agent 名称")
|
|
source_model = Column(String(100), nullable=True, comment="来源模型")
|
|
|
|
# RAG 检索
|
|
embedding_text = Column(Text, nullable=True, comment="用于生成 embedding 的合并文本")
|
|
embedding = Column(Text, nullable=True, comment="JSON 序列化的 embedding 向量")
|
|
|
|
# 效果度量
|
|
retrieval_count = Column(Integer, default=0, comment="被检索次数")
|
|
success_rate = Column(Float, nullable=True, comment="应用成功率")
|
|
|
|
# 提取信息
|
|
extracted_by = Column(String(100), nullable=True, comment="提取方式: llm_auto/manual/reviewed")
|
|
confidence = Column(Float, default=0.5, comment="提取置信度(0-1)")
|
|
|
|
is_active = Column(Boolean, default=True, comment="是否启用")
|
|
created_at = Column(DateTime, default=datetime.now, comment="创建时间")
|
|
updated_at = Column(DateTime, default=datetime.now, onupdate=datetime.now, comment="更新时间")
|
|
|
|
__table_args__ = (
|
|
Index("ix_knowledge_entries_category", "category"),
|
|
Index("ix_knowledge_entries_active", "is_active"),
|
|
)
|
|
|
|
def __repr__(self):
|
|
return f"<KnowledgeEntry(id={self.id}, title={self.title}, category={self.category})>"
|
|
|
|
def to_dict(self) -> dict:
|
|
return {
|
|
"id": self.id,
|
|
"title": self.title,
|
|
"category": self.category,
|
|
"tags": self.tags or [],
|
|
"situation": self.situation,
|
|
"solution": self.solution,
|
|
"caveats": self.caveats,
|
|
"source_agent_name": self.source_agent_name,
|
|
"retrieval_count": self.retrieval_count,
|
|
"confidence": self.confidence,
|
|
"created_at": self.created_at.isoformat() if self.created_at else None,
|
|
}
|