feat: Phase 1+2 架构进化 — 连续思考链/主动消息决策/情感状态机/离线自主思考 (86文件)

Phase 1 (基础设施):
- ThinkChain 思考链连续性 + 差异化思考提示词 (persistent)
- AutonomousToolPolicy 工具安全策略 (safe/unsafe/conditional)
- MessageScheduler 自适应消息节奏 (Idle/Available/Busy)
- SessionEnrichmentStore 渐进式上下文丰富 (5层)
- ConversationBus 事件总线 + ResponseCache (dedup)
- pkg/logger 统一日志 + 所有 handler 替换 fmt.Printf
- NPE 守卫/链路优化/数据库表修复/Go workspace

Phase 2 (人格交互):
- EmotionState/EmotionTracker 情感状态机 (5种心情, 情绪衰减)
- ProactiveGuard 主动消息多维决策 (静默时段/紧急度/频率/校验)
- Gateway↔ai-core 在线状态感知链路 (presence notification)
- 离线思考频率控制 + 重连问候 + 离线消息排队

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
2026-05-23 15:25:12 +08:00
parent b123a36aae
commit 87214b9441
86 changed files with 3085 additions and 582 deletions
@@ -3,7 +3,7 @@ package subsession
import (
"context"
"fmt"
"log"
"github.com/yourname/cyrene-ai/pkg/logger"
"time"
"github.com/yourname/cyrene-ai/ai-core/internal/llm"
@@ -29,8 +29,10 @@ func (p *GeneralProvider) Type() model.SubSessionType {
}
func (p *GeneralProvider) CanHandle(_ context.Context, _ *model.IntentResult, _ string) bool {
// General 子会话总是需要(核心对话逻辑)
return true
// Phase 1 Step 2: GeneralProvider is a no-op (Execute returns hardcoded string).
// Chat synthesis is handled directly by the orchestrator's Synthesizer.
// Disabled to avoid wasting a goroutine + LLM context creation.
return false
}
func (p *GeneralProvider) Priority() int {
@@ -123,7 +125,7 @@ func (p *GeneralProvider) Execute(ctx context.Context, subCtx []model.LLMMessage
// 由于 GeneralProvider 暂时不需要工具调用等特殊逻辑,我们返回一个简单的摘要标记,
// 实际的 LLM 调用将在 orchestrator 中完成(通过 Manager.Dispatch 后的 llmClient)。
log.Printf("[general-subsession] 通用对话子会话上下文已创建 (%d 条消息)", len(subCtx))
logger.Printf("[general-subsession] 通用对话子会话上下文已创建 (%d 条消息)", len(subCtx))
return &model.SubSessionResult{
Type: model.SubSessionGeneral,
Summary: "思考完成,等待主会话综合",