Version: 0.9.26.dev.260417
后端： 1. Prompt 层从 execute 专属骨架重构为全节点统一四段式 buildUnifiedStageMessages - 新增 unified_context.go：定义 StageMessagesConfig + buildUnifiedStageMessages 统一骨架，所有节点（Chat/Plan/Execute/Deliver/DeepAnswer）共用同一套 msg0~msg3 拼装逻辑 - 新增 conversation_view.go：通用对话历史渲染 buildConversationHistoryMessage，各节点复用，不再各自维护提取逻辑 - 新增 chat_context.go / plan_context.go / deliver_context.go：各节点自行渲染 msg1（对话视图）和 msg2（工作区），统一层只负责"怎么拼"，不再替节点决定"放什么" - Chat/Plan/Deliver/Execute 的 BuildXXXMessages 全部从 buildStageMessages 切到 buildUnifiedStageMessages，移除旧路径 - 删除 execute_pinned.go：execute 记忆渲染合并到统一层 renderUnifiedMemoryContext - Plan prompt 不再在 user prompt 中拼装任务类 ID 列表和 renderStateSummary，改为依赖 msg2 规划工作区；Chat 粗排判断从"上下文有任务类 ID"改为"批量调度需求" - Deliver prompt 新增 IsAborted/IsExhaustedTerminal 区分，支持粗排收口和主动终止场景 2. Execute ReAct 上下文简化——移除归档搬运、窗口裁剪和重复工具压缩 - 移除 splitExecuteLoopRecordsByBoundary、findLatestExecuteBoundaryMarker、tailExecuteLoops、compressExecuteLoopObservationsByTool、buildEarlyExecuteReactSummary、trimExecuteMessage1ByBudget 等六个函数 - 移除 executeLoopWindowLimit / executeConversationTurnLimit / executeMessage1MaxRunes 等预算常量 - msg1 不再从历史中归档上一轮 ReAct 结果，只保留真实对话流（user + assistant speak），全量注入 - msg2 不再按 loop_closed / step_advanced 边界切分"归档/活跃"，直接全量注入全部 ReAct Loop 记录 - token 预算由统一压缩层兜底，prompt 层不再做提前裁剪 3. 压缩层从 Execute 专属提升为全节点通用 UnifiedCompact - 删除 execute_compact.go（Execute 专属压缩文件） - 新增 unified_compact.go：UnifiedCompactInput 参数化，各节点（Plan/Chat/Deliver/Execute）构造时从自己的 NodeInput 提取公共字段，消除对 Execute 的直接依赖 - CompactionStore 接口扩展 LoadStageCompaction / SaveStageCompaction，各节点按 stageKey 独立维护压缩状态互不覆盖 - 非 4 段式消息时退化成按角色汇总统计，确保 context_token_stats 仍然刷新 4. Retry 重试机制全面下线 - dao/agent.go：saveChatHistoryCore / SaveChatHistory / SaveChatHistoryInTx 移除 retry_group_id / retry_index / retry_from_user_message_id / retry_from_assistant_message_id 四个参数，修复乱码注释 - dao/agent-cache.go：移除 ApplyRetrySeed 和 extractMessageHistoryID 两个方法 - conv/agent.go：ToEinoMessages 不再回灌 retry_* 字段到运行期上下文 - service/agentsvc/agent.go：移除 chatRetryMeta 及 resolveRetryGroupID / buildRetrySeed 等全部重试逻辑 - service/agentsvc/agent_quick_note.go：整个文件删除（retry 快速补写路径已无用） - service/events/chat_history_persist.go：移除 retry 参数传递 5. 节点层瘦身 + 可见消息逐条持久化 - agent_nodes.go 大幅简化：Chat/Plan/Execute/Deliver 节点方法移除 ToolSchema 注入、状态摘要渲染等逻辑，只做参数转发和状态落盘 - 新增 visible_message.go：persistVisibleAssistantMessage 统一处理可见 assistant speak 的实时持久化，失败仅记日志不中断主流程 - 新增 llm_debug.go：logNodeLLMContext 统一打印 LLM 上下文调试日志 - graph_run_state.go 新增 PersistVisibleMessageFunc 类型 + AgentGraphDeps.PersistVisibleMessage 字段 - service/agentsvc/agent_newagent.go 精简主循环，注入 PersistVisibleMessage 回调；agent_history.go 精简历史构建 - token_budget.go 移除 Execute 专属预算检查，统一到通用预算 前端： 1. 移除 retry 相关 UI 和类型 - agent.ts 移除 retry_group_id / retry_index / retry_total 字段及 normalize 逻辑 - AssistantPanel.vue 移除 retry 相关 UI 和交互代码（约 700 行精简） - dashboard.ts 移除 retry 相关类型定义 - AssistantView.vue 微调 2. ContextWindowMeter 压缩次数展示和数值格式优化 - 新增 formatCompactCount 工具函数，千位以上用 k 单位压缩（如 80k） - 新增压缩次数显示 3.修复了新对话发消息时，user和assistant消息被自动调换的bug 仓库：无
This commit is contained in:
Losita
@@ -23,23 +23,12 @@ func ToEinoMessages(dbMsgs []model.ChatHistory) []*schema.Message {
|
||||
Content: safeChatHistoryText(m.MessageContent),
|
||||
ReasoningContent: safeChatHistoryText(m.ReasoningContent),
|
||||
}
|
||||
// retry 机制已整体下线：历史数据里的 retry_* 列不再回灌到运行期上下文。
|
||||
extra := make(map[string]any)
|
||||
extra["history_id"] = m.ID
|
||||
if m.ReasoningDurationSeconds > 0 {
|
||||
extra["reasoning_duration_seconds"] = m.ReasoningDurationSeconds
|
||||
}
|
||||
if m.RetryGroupID != nil && *m.RetryGroupID != "" {
|
||||
extra["retry_group_id"] = *m.RetryGroupID
|
||||
}
|
||||
if m.RetryIndex != nil && *m.RetryIndex > 0 {
|
||||
extra["retry_index"] = *m.RetryIndex
|
||||
}
|
||||
if m.RetryFromUserMessageID != nil && *m.RetryFromUserMessageID > 0 {
|
||||
extra["retry_from_user_message_id"] = *m.RetryFromUserMessageID
|
||||
}
|
||||
if m.RetryFromAssistantMessageID != nil && *m.RetryFromAssistantMessageID > 0 {
|
||||
extra["retry_from_assistant_message_id"] = *m.RetryFromAssistantMessageID
|
||||
}
|
||||
if len(extra) > 0 {
|
||||
msg.Extra = extra
|
||||
}
|
||||
|
||||
@@ -5,7 +5,6 @@ import (
|
||||
"encoding/json"
|
||||
"fmt"
|
||||
"strconv"
|
||||
"strings"
|
||||
"time"
|
||||
|
||||
"github.com/cloudwego/eino/schema"
|
||||
@@ -164,79 +163,6 @@ func (m *AgentCache) BackfillHistory(ctx context.Context, sessionID string, mess
|
||||
return err
|
||||
}
|
||||
|
||||
func (m *AgentCache) ApplyRetrySeed(ctx context.Context, sessionID, retryGroupID string, sourceUserMessageID, sourceAssistantMessageID int) error {
|
||||
if m == nil || m.client == nil {
|
||||
return nil
|
||||
}
|
||||
groupID := strings.TrimSpace(retryGroupID)
|
||||
if groupID == "" {
|
||||
return nil
|
||||
}
|
||||
|
||||
vals, err := m.client.LRange(ctx, m.historyKey(sessionID), 0, -1).Result()
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
if len(vals) == 0 {
|
||||
return nil
|
||||
}
|
||||
|
||||
changed := false
|
||||
targets := map[int]struct{}{}
|
||||
if sourceUserMessageID > 0 {
|
||||
targets[sourceUserMessageID] = struct{}{}
|
||||
}
|
||||
if sourceAssistantMessageID > 0 {
|
||||
targets[sourceAssistantMessageID] = struct{}{}
|
||||
}
|
||||
if len(targets) == 0 {
|
||||
return nil
|
||||
}
|
||||
|
||||
indexOne := 1
|
||||
for idx, raw := range vals {
|
||||
var msg schema.Message
|
||||
if err := json.Unmarshal([]byte(raw), &msg); err != nil {
|
||||
return err
|
||||
}
|
||||
historyID := extractMessageHistoryID(&msg)
|
||||
if historyID <= 0 {
|
||||
continue
|
||||
}
|
||||
if _, ok := targets[historyID]; !ok {
|
||||
continue
|
||||
}
|
||||
if msg.Extra == nil {
|
||||
msg.Extra = make(map[string]any)
|
||||
}
|
||||
msg.Extra["retry_group_id"] = groupID
|
||||
msg.Extra["retry_index"] = indexOne
|
||||
updated, err := json.Marshal(&msg)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
vals[idx] = string(updated)
|
||||
changed = true
|
||||
}
|
||||
|
||||
if !changed {
|
||||
return nil
|
||||
}
|
||||
|
||||
pipe := m.client.Pipeline()
|
||||
key := m.historyKey(sessionID)
|
||||
pipe.Del(ctx, key)
|
||||
values := make([]interface{}, 0, len(vals))
|
||||
for _, item := range vals {
|
||||
values = append(values, item)
|
||||
}
|
||||
pipe.RPush(ctx, key, values...)
|
||||
pipe.LTrim(ctx, key, 0, int64(len(vals)-1))
|
||||
pipe.Expire(ctx, key, m.expiration)
|
||||
_, err = pipe.Exec(ctx)
|
||||
return err
|
||||
}
|
||||
|
||||
func (m *AgentCache) ClearHistory(ctx context.Context, sessionID string) error {
|
||||
historyKey := m.historyKey(sessionID)
|
||||
windowKey := m.historyWindowKey(sessionID)
|
||||
@@ -263,49 +189,6 @@ func (m *AgentCache) DeleteConversationStatus(ctx context.Context, sessionID str
|
||||
return m.client.Del(ctx, key).Err()
|
||||
}
|
||||
|
||||
func extractMessageHistoryID(msg *schema.Message) int {
|
||||
if msg == nil || msg.Extra == nil {
|
||||
return 0
|
||||
}
|
||||
raw, ok := msg.Extra["history_id"]
|
||||
if !ok {
|
||||
return 0
|
||||
}
|
||||
// 1. history_id 主要来自 DB 回填，正常情况下是 number。
|
||||
// 2. 但 Redis 往返、灰度期数据修复或手工写入时，仍可能出现字符串数字。
|
||||
// 3. 这里做一次宽松解析，避免重试分组补种时因为类型差异找不到源消息。
|
||||
switch v := raw.(type) {
|
||||
case int:
|
||||
return v
|
||||
case int32:
|
||||
return int(v)
|
||||
case int64:
|
||||
return int(v)
|
||||
case float64:
|
||||
return int(v)
|
||||
case json.Number:
|
||||
if parsed, err := v.Int64(); err == nil {
|
||||
return int(parsed)
|
||||
}
|
||||
if parsed, err := v.Float64(); err == nil {
|
||||
return int(parsed)
|
||||
}
|
||||
return 0
|
||||
case string:
|
||||
trimmed := strings.TrimSpace(v)
|
||||
if trimmed == "" {
|
||||
return 0
|
||||
}
|
||||
parsed, err := strconv.Atoi(trimmed)
|
||||
if err != nil {
|
||||
return 0
|
||||
}
|
||||
return parsed
|
||||
default:
|
||||
return 0
|
||||
}
|
||||
}
|
||||
|
||||
// ---- Compaction 缓存 ----
|
||||
|
||||
func (m *AgentCache) compactionKey(chatID string) string {
|
||||
|
||||
@@ -2,6 +2,7 @@ package dao
|
||||
|
||||
import (
|
||||
"context"
|
||||
"encoding/json"
|
||||
"errors"
|
||||
"fmt"
|
||||
"strings"
|
||||
@@ -23,18 +24,23 @@ func (r *AgentDAO) WithTx(tx *gorm.DB) *AgentDAO {
|
||||
return &AgentDAO{db: tx}
|
||||
}
|
||||
|
||||
// saveChatHistoryCore 鏄€滆亰澶╂秷鎭惤搴?+ 浼氳瘽缁熻鏇存柊鈥濈殑鏍稿績瀹炵幇銆?
|
||||
// saveChatHistoryCore 是"聊天消息落库 + 会话统计更新"的核心实现。
|
||||
//
|
||||
// 鑱岃矗杈圭晫锛?
|
||||
// 1. 鍙墽琛屽綋鍓?DAO 鍙ユ焺涓婄殑鏁版嵁搴撳啓鍏ュ姩浣滐紱
|
||||
// 2. 涓嶄富鍔ㄥ紑鍚簨鍔★紙浜嬪姟鐢辫皟鐢ㄦ柟鍐冲畾锛夛紱
|
||||
// 3. 淇濊瘉 chat_histories 涓?agent_chats.message_count 鐨勪竴鑷存€у彛寰勩€?
|
||||
// 职责边界：
|
||||
// 1. 只执行当前 DAO 句柄上的数据库写入动作；
|
||||
// 2. 不主动开启事务（事务由调用方决定）；
|
||||
// 3. 保证 chat_histories 与 agent_chats.message_count 的一致性口径。
|
||||
//
|
||||
// 澶辫触澶勭悊锛?
|
||||
// 1. 浠讳竴姝ラ澶辫触閮借繑鍥?error锛?
|
||||
// 2. 鑻ヨ皟鐢ㄦ柟澶勪簬浜嬪姟涓紝杩斿洖 error 浼氳Е鍙戜簨鍔″洖婊氥€?
|
||||
func (a *AgentDAO) saveChatHistoryCore(ctx context.Context, userID int, conversationID string, role, message, reasoningContent string, reasoningDurationSeconds int, retryGroupID *string, retryIndex *int, retryFromUserMessageID *int, retryFromAssistantMessageID *int, tokensConsumed int) error {
|
||||
// 0. token 鍏ュ簱鍓嶅厹搴曪細璐熸暟缁熶竴褰掗浂锛岄伩鍏嶅紓甯稿€兼薄鏌撶疮璁＄粺璁°€?
|
||||
// 失败处理：
|
||||
// 1. 任一步骤失败都返回 error；
|
||||
// 2. 若调用方处于事务中，返回 error 会触发事务回滚。
|
||||
//
|
||||
// 关于 retry 字段：
|
||||
// 1. retry 机制已整体下线，本函数不再写入 retry_group_id / retry_index / retry_from_* 四列；
|
||||
// 2. 这些列在 GORM ChatHistory 模型上暂时保留，列本身可空，历史数据不受影响；
|
||||
// 3. Step B 会做 DROP COLUMN 的 migration。
|
||||
func (a *AgentDAO) saveChatHistoryCore(ctx context.Context, userID int, conversationID string, role, message, reasoningContent string, reasoningDurationSeconds int, tokensConsumed int) error {
|
||||
// 0. token 入库前兜底：负数统一归零，避免异常值污染累计统计。
|
||||
if tokensConsumed < 0 {
|
||||
tokensConsumed = 0
|
||||
}
|
||||
@@ -43,32 +49,28 @@ func (a *AgentDAO) saveChatHistoryCore(ctx context.Context, userID int, conversa
|
||||
reasoningDurationSeconds = 0
|
||||
}
|
||||
|
||||
// 1. 鍏堝啓 chat_histories 鍘熷娑堟伅銆?
|
||||
// 1. 先写 chat_histories 原始消息。
|
||||
var reasoningContentPtr *string
|
||||
if reasoningContent != "" {
|
||||
reasoningContentPtr = &reasoningContent
|
||||
}
|
||||
userChat := model.ChatHistory{
|
||||
UserID: userID,
|
||||
MessageContent: &message,
|
||||
ReasoningContent: reasoningContentPtr,
|
||||
ReasoningDurationSeconds: reasoningDurationSeconds,
|
||||
RetryGroupID: retryGroupID,
|
||||
RetryIndex: retryIndex,
|
||||
RetryFromUserMessageID: retryFromUserMessageID,
|
||||
RetryFromAssistantMessageID: retryFromAssistantMessageID,
|
||||
Role: &role,
|
||||
ChatID: conversationID,
|
||||
TokensConsumed: tokensConsumed,
|
||||
UserID: userID,
|
||||
MessageContent: &message,
|
||||
ReasoningContent: reasoningContentPtr,
|
||||
ReasoningDurationSeconds: reasoningDurationSeconds,
|
||||
Role: &role,
|
||||
ChatID: conversationID,
|
||||
TokensConsumed: tokensConsumed,
|
||||
}
|
||||
if err := a.db.WithContext(ctx).Create(&userChat).Error; err != nil {
|
||||
return err
|
||||
}
|
||||
|
||||
// 2. 鍐嶆洿鏂颁細璇濈粺璁★細
|
||||
// 2.1 message_count +1锛屼繚鎸佸拰 chat_histories 琛屾暟鍙ｅ緞涓€鑷达紱
|
||||
// 2.2 tokens_total 绱姞鏈潯娑堟伅 token锛?
|
||||
// 2.3 last_message_at 鍒锋柊涓哄綋鍓嶆椂闂达紝渚涗細璇濇帓搴忎娇鐢ㄣ€?
|
||||
// 2. 再更新会话统计：
|
||||
// 2.1 message_count +1，保持和 chat_histories 行数口径一致；
|
||||
// 2.2 tokens_total 累加本条消息 token；
|
||||
// 2.3 last_message_at 刷新为当前时间，供会话排序使用。
|
||||
now := time.Now()
|
||||
updates := map[string]interface{}{
|
||||
"message_count": gorm.Expr("message_count + ?", 1),
|
||||
@@ -82,14 +84,14 @@ func (a *AgentDAO) saveChatHistoryCore(ctx context.Context, userID int, conversa
|
||||
return result.Error
|
||||
}
|
||||
if result.RowsAffected == 0 {
|
||||
// 浼氳瘽涓嶅瓨鍦ㄦ椂鐩存帴澶辫触锛岄伩鍏嶅嚭鐜扳€滃鍎垮巻鍙叉秷鎭€濄€?
|
||||
// 会话不存在时直接失败，避免出现"孤儿历史消息"。
|
||||
return fmt.Errorf("conversation not found when updating stats: user_id=%d chat_id=%s", userID, conversationID)
|
||||
}
|
||||
|
||||
// 3. 鏈€鍚庢洿鏂?users.token_usage锛堝悓涓€浜嬪姟鍐咃級锛?
|
||||
// 3.1 鍙湪 tokensConsumed>0 鏃舵墽琛岋紝閬垮厤鏃犳剰涔夊啓鍏ワ紱
|
||||
// 3.2 鍜?chat_histories/agent_chats 鏀惧湪鍚屼竴浜嬪姟閲岋紝淇濊瘉缁熻鍙ｅ緞鍘熷瓙涓€鑷达紱
|
||||
// 3.3 鑻ョ敤鎴疯涓嶅瓨鍦ㄥ垯杩斿洖閿欒锛岃Е鍙戜簨鍔″洖婊氾紝闃叉鍑虹幇鈥滀細璇濈粺璁℃垚鍔熶絾鐢ㄦ埛缁熻涓㈠け鈥濄€?
|
||||
// 3. 最后更新 users.token_usage（同一事务内）：
|
||||
// 3.1 只在 tokensConsumed>0 时执行，避免无意义写入；
|
||||
// 3.2 和 chat_histories/agent_chats 放在同一事务里，保证统计口径原子一致；
|
||||
// 3.3 若用户行不存在则返回错误，触发事务回滚，防止出现"会话统计成功但用户统计丢失"。
|
||||
if tokensConsumed > 0 {
|
||||
userUpdate := a.db.WithContext(ctx).
|
||||
Model(&model.User{}).
|
||||
@@ -105,38 +107,38 @@ func (a *AgentDAO) saveChatHistoryCore(ctx context.Context, userID int, conversa
|
||||
return nil
|
||||
}
|
||||
|
||||
// SaveChatHistoryInTx 鍦ㄨ皟鐢ㄦ柟鈥滃凡寮€鍚簨鍔♀€濈殑鍦烘櫙涓嬪啓鍏ヨ亰澶╁巻鍙层€?
|
||||
// SaveChatHistoryInTx 在调用方"已开启事务"的场景下写入聊天历史。
|
||||
//
|
||||
// 璁捐鐩殑锛?
|
||||
// 1. 缁欐湇鍔″眰缁勫悎澶氫釜 DAO 鎿嶄綔鏃跺鐢紝閬垮厤宓屽浜嬪姟锛?
|
||||
// 2. 璁?outbox 娑堣垂澶勭悊鍣ㄥ彲浠ュ拰涓氬姟鍐欏叆鍏变韩鍚屼竴涓?tx銆?
|
||||
func (a *AgentDAO) SaveChatHistoryInTx(ctx context.Context, userID int, conversationID string, role, message, reasoningContent string, reasoningDurationSeconds int, retryGroupID *string, retryIndex *int, retryFromUserMessageID *int, retryFromAssistantMessageID *int, tokensConsumed int) error {
|
||||
return a.saveChatHistoryCore(ctx, userID, conversationID, role, message, reasoningContent, reasoningDurationSeconds, retryGroupID, retryIndex, retryFromUserMessageID, retryFromAssistantMessageID, tokensConsumed)
|
||||
// 设计目的：
|
||||
// 1. 给服务层组合多个 DAO 操作时复用，避免嵌套事务；
|
||||
// 2. 让 outbox 消费处理器可以和业务写入共享同一个 tx。
|
||||
func (a *AgentDAO) SaveChatHistoryInTx(ctx context.Context, userID int, conversationID string, role, message, reasoningContent string, reasoningDurationSeconds int, tokensConsumed int) error {
|
||||
return a.saveChatHistoryCore(ctx, userID, conversationID, role, message, reasoningContent, reasoningDurationSeconds, tokensConsumed)
|
||||
}
|
||||
|
||||
// SaveChatHistory 鍦ㄥ悓姝ョ洿鍐欒矾寰勪笅鍐欏叆鑱婂ぉ鍘嗗彶銆?
|
||||
// SaveChatHistory 在同步直写路径下写入聊天历史。
|
||||
//
|
||||
// 璇存槑锛?
|
||||
// 1. 璇ユ柟娉曚細鑷寮€鍚簨鍔★紱
|
||||
// 2. 鍐呴儴澶嶇敤 saveChatHistoryCore锛岀‘淇濆拰 SaveChatHistoryInTx 鐨勪笟鍔″彛寰勫畬鍏ㄤ竴鑷淬€?
|
||||
func (a *AgentDAO) SaveChatHistory(ctx context.Context, userID int, conversationID string, role, message, reasoningContent string, reasoningDurationSeconds int, retryGroupID *string, retryIndex *int, retryFromUserMessageID *int, retryFromAssistantMessageID *int, tokensConsumed int) error {
|
||||
// 说明：
|
||||
// 1. 该方法会自行开启事务；
|
||||
// 2. 内部复用 saveChatHistoryCore，确保和 SaveChatHistoryInTx 的业务口径完全一致。
|
||||
func (a *AgentDAO) SaveChatHistory(ctx context.Context, userID int, conversationID string, role, message, reasoningContent string, reasoningDurationSeconds int, tokensConsumed int) error {
|
||||
return a.db.WithContext(ctx).Transaction(func(tx *gorm.DB) error {
|
||||
return a.WithTx(tx).saveChatHistoryCore(ctx, userID, conversationID, role, message, reasoningContent, reasoningDurationSeconds, retryGroupID, retryIndex, retryFromUserMessageID, retryFromAssistantMessageID, tokensConsumed)
|
||||
return a.WithTx(tx).saveChatHistoryCore(ctx, userID, conversationID, role, message, reasoningContent, reasoningDurationSeconds, tokensConsumed)
|
||||
})
|
||||
}
|
||||
|
||||
// adjustTokenUsageCore 鍦ㄥ悓涓€浜嬪姟璇箟涓嬪仛鈥滀細璇?鐢ㄦ埛鈥漷oken 璐︽湰澧為噺璋冩暣銆?
|
||||
// adjustTokenUsageCore 在同一事务语义下做"会话/用户"token 账本增量调整。
|
||||
//
|
||||
// 鑱岃矗杈圭晫锛?
|
||||
// 1. 鍙洿鏂?agent_chats.tokens_total 涓?users.token_usage锛?
|
||||
// 2. 涓嶅啓 chat_histories锛堟秷鎭惤搴撶敱 SaveChatHistory* 璺緞璐熻矗锛夛紱
|
||||
// 3. deltaTokens<=0 鏃惰涓烘棤鎿嶄綔锛岀洿鎺ヨ繑鍥炪€?
|
||||
// 职责边界：
|
||||
// 1. 只更新 agent_chats.tokens_total 与 users.token_usage；
|
||||
// 2. 不写 chat_histories（消息落库由 SaveChatHistory* 路径负责）；
|
||||
// 3. deltaTokens<=0 时视为无操作，直接返回。
|
||||
func (a *AgentDAO) adjustTokenUsageCore(ctx context.Context, userID int, conversationID string, deltaTokens int) error {
|
||||
if deltaTokens <= 0 {
|
||||
return nil
|
||||
}
|
||||
|
||||
// 1. 鍏堟洿鏂颁細璇濈疮璁?token銆?
|
||||
// 1. 先更新会话累计 token。
|
||||
chatUpdate := a.db.WithContext(ctx).
|
||||
Model(&model.AgentChat{}).
|
||||
Where("user_id = ? AND chat_id = ?", userID, conversationID).
|
||||
@@ -148,7 +150,7 @@ func (a *AgentDAO) adjustTokenUsageCore(ctx context.Context, userID int, convers
|
||||
return fmt.Errorf("conversation not found when adjusting tokens: user_id=%d chat_id=%s", userID, conversationID)
|
||||
}
|
||||
|
||||
// 2. 鍐嶆洿鏂扮敤鎴风疮璁?token銆?
|
||||
// 2. 再更新用户累计 token。
|
||||
userUpdate := a.db.WithContext(ctx).
|
||||
Model(&model.User{}).
|
||||
Where("id = ?", userID).
|
||||
@@ -162,12 +164,12 @@ func (a *AgentDAO) adjustTokenUsageCore(ctx context.Context, userID int, convers
|
||||
return nil
|
||||
}
|
||||
|
||||
// AdjustTokenUsageInTx 鍦ㄨ皟鐢ㄦ柟宸插紑鍚簨鍔℃椂鎵ц token 璐︽湰澧為噺璋冩暣銆?
|
||||
// AdjustTokenUsageInTx 在调用方已开启事务时执行 token 账本增量调整。
|
||||
func (a *AgentDAO) AdjustTokenUsageInTx(ctx context.Context, userID int, conversationID string, deltaTokens int) error {
|
||||
return a.adjustTokenUsageCore(ctx, userID, conversationID, deltaTokens)
|
||||
}
|
||||
|
||||
// AdjustTokenUsage 鍦ㄥ悓姝ヨ矾寰勪笅鎵ц token 璐︽湰澧為噺璋冩暣锛堝唴閮ㄨ嚜甯︿簨鍔★級銆?
|
||||
// AdjustTokenUsage 在同步路径下执行 token 账本增量调整（内部自带事务）。
|
||||
func (a *AgentDAO) AdjustTokenUsage(ctx context.Context, userID int, conversationID string, deltaTokens int) error {
|
||||
return a.db.WithContext(ctx).Transaction(func(tx *gorm.DB) error {
|
||||
return a.WithTx(tx).adjustTokenUsageCore(ctx, userID, conversationID, deltaTokens)
|
||||
@@ -197,110 +199,13 @@ func (a *AgentDAO) GetUserChatHistories(ctx context.Context, userID, limit int,
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
// 淇濈暀鈥滄渶杩?N 鏉♀€濆悗锛屽弽杞垚鏃堕棿姝ｅ簭锛屾柟渚挎ā鍨嬫秷璐广€?
|
||||
// 保留"最近 N 条"后，反转成时间正序，方便模型消费。
|
||||
for i, j := 0, len(histories)-1; i < j; i, j = i+1, j-1 {
|
||||
histories[i], histories[j] = histories[j], histories[i]
|
||||
}
|
||||
return histories, nil
|
||||
}
|
||||
|
||||
func (a *AgentDAO) EnsureRetryGroupSeed(ctx context.Context, userID int, chatID, retryGroupID string, sourceUserMessageID, sourceAssistantMessageID int) error {
|
||||
normalizedGroupID := strings.TrimSpace(retryGroupID)
|
||||
if normalizedGroupID == "" {
|
||||
return nil
|
||||
}
|
||||
|
||||
indexOne := 1
|
||||
ids := make([]int, 0, 2)
|
||||
if sourceUserMessageID > 0 {
|
||||
ids = append(ids, sourceUserMessageID)
|
||||
}
|
||||
if sourceAssistantMessageID > 0 {
|
||||
ids = append(ids, sourceAssistantMessageID)
|
||||
}
|
||||
if len(ids) == 0 {
|
||||
return nil
|
||||
}
|
||||
|
||||
return a.db.WithContext(ctx).
|
||||
Model(&model.ChatHistory{
|
||||
UserID: userID,
|
||||
ChatID: chatID,
|
||||
}).
|
||||
Where("user_id = ? AND chat_id = ? AND id IN ?", userID, chatID, ids).
|
||||
Where("(retry_group_id IS NULL OR retry_group_id = '')").
|
||||
Updates(map[string]any{
|
||||
"retry_group_id": normalizedGroupID,
|
||||
"retry_index": indexOne,
|
||||
}).Error
|
||||
}
|
||||
|
||||
// ValidateRetrySourceMessages 校验重试父消息是否真实存在且角色匹配。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 负责校验 retry 请求引用的父 user/assistant 消息是否属于当前用户、当前会话。
|
||||
// 2. 负责校验两条父消息的角色语义，避免把占位 id、串号 id 或交换角色的 id 写进数据库。
|
||||
// 3. 不负责补种 retry_group_id；分组补种仍由 EnsureRetryGroupSeed 负责。
|
||||
func (a *AgentDAO) ValidateRetrySourceMessages(ctx context.Context, userID int, chatID string, sourceUserMessageID, sourceAssistantMessageID int) error {
|
||||
// 1. retry 是“基于既有一问一答重新生成”，因此两条父消息 id 必须同时有效。
|
||||
// 2. 只要任意一个缺失，就直接返回错误，禁止继续写出 index=1 的脏重试数据。
|
||||
if sourceUserMessageID <= 0 || sourceAssistantMessageID <= 0 {
|
||||
return errors.New("retry source message ids are invalid")
|
||||
}
|
||||
|
||||
type retrySourceRow struct {
|
||||
ID int
|
||||
Role *string
|
||||
}
|
||||
|
||||
ids := []int{sourceUserMessageID, sourceAssistantMessageID}
|
||||
rows := make([]retrySourceRow, 0, len(ids))
|
||||
if err := a.db.WithContext(ctx).
|
||||
Model(&model.ChatHistory{}).
|
||||
Select("id", "role").
|
||||
Where("user_id = ? AND chat_id = ? AND id IN ?", userID, chatID, ids).
|
||||
Find(&rows).Error; err != nil {
|
||||
return err
|
||||
}
|
||||
if len(rows) != len(ids) {
|
||||
return errors.New("retry source messages not found in current conversation")
|
||||
}
|
||||
|
||||
roleByID := make(map[int]string, len(rows))
|
||||
for _, row := range rows {
|
||||
if row.Role == nil {
|
||||
roleByID[row.ID] = ""
|
||||
continue
|
||||
}
|
||||
roleByID[row.ID] = strings.ToLower(strings.TrimSpace(*row.Role))
|
||||
}
|
||||
|
||||
if roleByID[sourceUserMessageID] != "user" {
|
||||
return errors.New("retry source user message is invalid")
|
||||
}
|
||||
if roleByID[sourceAssistantMessageID] != "assistant" {
|
||||
return errors.New("retry source assistant message is invalid")
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func (a *AgentDAO) GetRetryGroupNextIndex(ctx context.Context, userID int, chatID, retryGroupID string) (int, error) {
|
||||
normalizedGroupID := strings.TrimSpace(retryGroupID)
|
||||
if normalizedGroupID == "" {
|
||||
return 0, errors.New("retry_group_id is empty")
|
||||
}
|
||||
|
||||
var maxIndex int
|
||||
if err := a.db.WithContext(ctx).
|
||||
Model(&model.ChatHistory{}).
|
||||
Where("user_id = ? AND chat_id = ? AND retry_group_id = ?", userID, chatID, normalizedGroupID).
|
||||
Select("COALESCE(MAX(retry_index), 0)").
|
||||
Scan(&maxIndex).Error; err != nil {
|
||||
return 0, err
|
||||
}
|
||||
return maxIndex + 1, nil
|
||||
}
|
||||
|
||||
func (a *AgentDAO) IfChatExists(ctx context.Context, userID int, chatID string) (bool, error) {
|
||||
var chat model.AgentChat
|
||||
err := a.db.WithContext(ctx).Where("user_id = ? AND chat_id = ?", userID, chatID).First(&chat).Error
|
||||
@@ -313,7 +218,7 @@ func (a *AgentDAO) IfChatExists(ctx context.Context, userID int, chatID string)
|
||||
return true, nil
|
||||
}
|
||||
|
||||
// GetConversationMeta 鏌ヨ鍗曚釜浼氳瘽鍏冧俊鎭€?
|
||||
// GetConversationMeta 查询单个会话元信息。
|
||||
func (a *AgentDAO) GetConversationMeta(ctx context.Context, userID int, chatID string) (*model.AgentChat, error) {
|
||||
var chat model.AgentChat
|
||||
err := a.db.WithContext(ctx).
|
||||
@@ -326,7 +231,7 @@ func (a *AgentDAO) GetConversationMeta(ctx context.Context, userID int, chatID s
|
||||
return &chat, nil
|
||||
}
|
||||
|
||||
// GetConversationTitle 璇诲彇褰撳墠浼氳瘽鏍囬銆?
|
||||
// GetConversationTitle 读取当前会话标题。
|
||||
func (a *AgentDAO) GetConversationTitle(ctx context.Context, userID int, chatID string) (title string, exists bool, err error) {
|
||||
var chat model.AgentChat
|
||||
queryErr := a.db.WithContext(ctx).
|
||||
@@ -345,7 +250,7 @@ func (a *AgentDAO) GetConversationTitle(ctx context.Context, userID int, chatID
|
||||
return strings.TrimSpace(*chat.Title), true, nil
|
||||
}
|
||||
|
||||
// UpdateConversationTitleIfEmpty 浠呭湪鏍囬涓虹┖鏃舵洿鏂颁細璇濇爣棰樸€?
|
||||
// UpdateConversationTitleIfEmpty 仅在标题为空时更新会话标题。
|
||||
func (a *AgentDAO) UpdateConversationTitleIfEmpty(ctx context.Context, userID int, chatID, title string) error {
|
||||
normalized := strings.TrimSpace(title)
|
||||
if normalized == "" {
|
||||
@@ -357,20 +262,20 @@ func (a *AgentDAO) UpdateConversationTitleIfEmpty(ctx context.Context, userID in
|
||||
Update("title", normalized).Error
|
||||
}
|
||||
|
||||
// GetConversationList 鎸夊垎椤垫煡璇㈡寚瀹氱敤鎴风殑浼氳瘽鍒楄〃銆?
|
||||
// GetConversationList 按分页查询指定用户的会话列表。
|
||||
//
|
||||
// 鑱岃矗杈圭晫锛?
|
||||
// 1. 鍙礋璐ｈ搴擄紝涓嶈礋璐ｇ紦瀛橈紱
|
||||
// 2. 鍙礋璐?user_id 鏁版嵁闅旂锛屼笉璐熻矗鍙傛暟鍚堟硶鎬у厹搴曪紙鐢?service 璐熻矗锛夛紱
|
||||
// 3. 杩斿洖鎬绘暟 total 渚涗笂灞傝绠?has_more銆?
|
||||
// 职责边界：
|
||||
// 1. 只负责读库，不负责缓存；
|
||||
// 2. 只负责 user_id 数据隔离，不负责参数合法性兜底（由 service 负责）；
|
||||
// 3. 返回总数 total 供上层计算 has_more。
|
||||
func (a *AgentDAO) GetConversationList(ctx context.Context, userID, page, pageSize int, status string) ([]model.AgentChat, int64, error) {
|
||||
// 1. 鍏堟瀯閫犵粺涓€杩囨护鏉′欢锛屼繚璇?total 涓?list 鐨勭粺璁″彛寰勪竴鑷淬€?
|
||||
// 1. 先构造统一过滤条件，保证 total 与 list 的统计口径一致。
|
||||
baseQuery := a.db.WithContext(ctx).Model(&model.AgentChat{}).Where("user_id = ?", userID)
|
||||
if strings.TrimSpace(status) != "" {
|
||||
baseQuery = baseQuery.Where("status = ?", status)
|
||||
}
|
||||
|
||||
// 2. 鍏堟煡鎬绘潯鏁帮紝缁欏墠绔垎椤靛櫒鎻愪緵瀹屾暣鍏冧俊鎭€?
|
||||
// 2. 先查总条数，给前端分页器提供完整元信息。
|
||||
var total int64
|
||||
if err := baseQuery.Count(&total).Error; err != nil {
|
||||
return nil, 0, err
|
||||
@@ -379,9 +284,9 @@ func (a *AgentDAO) GetConversationList(ctx context.Context, userID, page, pageSi
|
||||
return make([]model.AgentChat, 0), 0, nil
|
||||
}
|
||||
|
||||
// 3. 鍐嶆煡褰撳墠椤垫暟鎹細
|
||||
// 3.1 鎸夋渶杩戞秷鎭椂闂村€掑簭锛屼繚璇佲€滄渶杩戞椿璺冣€濅紭鍏堝睍绀猴紱
|
||||
// 3.2 鍚屾椂闂存埑涓嬫寜 id 鍊掑簭锛岄伩鍏嶇炕椤垫椂椤哄簭鎶栧姩銆?
|
||||
// 3. 再查当前页数据：
|
||||
// 3.1 按最近消息时间倒序，保证"最近活跃"优先展示；
|
||||
// 3.2 同时间戳下按 id 倒序，避免翻页时顺序抖动。
|
||||
offset := (page - 1) * pageSize
|
||||
var chats []model.AgentChat
|
||||
query := a.db.WithContext(ctx).
|
||||
@@ -402,34 +307,17 @@ func (a *AgentDAO) GetConversationList(ctx context.Context, userID, page, pageSi
|
||||
return chats, total, nil
|
||||
}
|
||||
|
||||
// ---- Compaction 相关 ----
|
||||
|
||||
// SaveCompaction 保存压缩摘要和水位线。
|
||||
// ---- 压缩摘要持久化 ----
|
||||
//
|
||||
// 1. 旧接口 SaveCompaction / LoadCompaction 继续保留，默认只读写 execute 阶段。
|
||||
// 2. 新接口按 stageKey 分桶读写，数据仍然落在 agent_chats.compaction_summary。
|
||||
// 3. 为兼容历史数据，若 compaction_summary 仍是旧字符串格式，则自动回退读取。
|
||||
func (a *AgentDAO) SaveCompaction(ctx context.Context, userID int, chatID string, summary string, watermark int) error {
|
||||
return a.db.WithContext(ctx).
|
||||
Model(&model.AgentChat{}).
|
||||
Where("user_id = ? AND chat_id = ?", userID, chatID).
|
||||
Updates(map[string]any{
|
||||
"compaction_summary": summary,
|
||||
"compaction_watermark": watermark,
|
||||
}).Error
|
||||
return a.SaveStageCompaction(ctx, userID, chatID, "execute", summary, watermark)
|
||||
}
|
||||
|
||||
// LoadCompaction 读取压缩摘要和水位线。
|
||||
func (a *AgentDAO) LoadCompaction(ctx context.Context, userID int, chatID string) (summary string, watermark int, err error) {
|
||||
var chat model.AgentChat
|
||||
err = a.db.WithContext(ctx).
|
||||
Select("compaction_summary", "compaction_watermark").
|
||||
Where("user_id = ? AND chat_id = ?", userID, chatID).
|
||||
First(&chat).Error
|
||||
if err != nil {
|
||||
return "", 0, err
|
||||
}
|
||||
if chat.CompactionSummary != nil {
|
||||
summary = *chat.CompactionSummary
|
||||
}
|
||||
watermark = chat.CompactionWatermark
|
||||
return
|
||||
return a.LoadStageCompaction(ctx, userID, chatID, "execute")
|
||||
}
|
||||
|
||||
// SaveContextTokenStats 保存上下文窗口 token 分布统计。
|
||||
@@ -455,3 +343,132 @@ func (a *AgentDAO) LoadContextTokenStats(ctx context.Context, userID int, chatID
|
||||
}
|
||||
return "", nil
|
||||
}
|
||||
|
||||
type stageCompactionRecord struct {
|
||||
Summary string `json:"summary"`
|
||||
Watermark int `json:"watermark"`
|
||||
}
|
||||
|
||||
type stageCompactionEnvelope struct {
|
||||
Version int `json:"version"`
|
||||
Stages map[string]stageCompactionRecord `json:"stages"`
|
||||
}
|
||||
|
||||
// normalizeCompactionStageKey 统一 stageKey 的写法，避免 "Execute" 和 "execute" 被当成两个键。
|
||||
func normalizeCompactionStageKey(stageKey string) string {
|
||||
key := strings.ToLower(strings.TrimSpace(stageKey))
|
||||
if key == "" {
|
||||
return "execute"
|
||||
}
|
||||
return key
|
||||
}
|
||||
|
||||
// loadStageCompactionStages 负责把数据库里的压缩摘要统一解包成 stage -> record。
|
||||
//
|
||||
// 1. 先处理空值，避免后续逻辑误判。
|
||||
// 2. 如果已经是 JSON envelope，就按 stage 逐项读取。
|
||||
// 3. 如果还是旧版纯字符串，就把它当作 execute 阶段的兼容数据。
|
||||
func loadStageCompactionStages(summary *string, watermark int) map[string]stageCompactionRecord {
|
||||
stages := map[string]stageCompactionRecord{}
|
||||
if summary == nil {
|
||||
return stages
|
||||
}
|
||||
|
||||
raw := strings.TrimSpace(*summary)
|
||||
if raw == "" {
|
||||
return stages
|
||||
}
|
||||
|
||||
var env stageCompactionEnvelope
|
||||
if err := json.Unmarshal([]byte(raw), &env); err == nil && len(env.Stages) > 0 {
|
||||
for key, record := range env.Stages {
|
||||
stages[normalizeCompactionStageKey(key)] = stageCompactionRecord{
|
||||
Summary: strings.TrimSpace(record.Summary),
|
||||
Watermark: record.Watermark,
|
||||
}
|
||||
}
|
||||
return stages
|
||||
}
|
||||
|
||||
stages["execute"] = stageCompactionRecord{
|
||||
Summary: raw,
|
||||
Watermark: watermark,
|
||||
}
|
||||
return stages
|
||||
}
|
||||
|
||||
// marshalStageCompactionStages 负责把按阶段分桶后的摘要重新编码为 JSON envelope。
|
||||
func marshalStageCompactionStages(stages map[string]stageCompactionRecord) (string, error) {
|
||||
env := stageCompactionEnvelope{
|
||||
Version: 1,
|
||||
Stages: stages,
|
||||
}
|
||||
data, err := json.Marshal(env)
|
||||
if err != nil {
|
||||
return "", err
|
||||
}
|
||||
return string(data), nil
|
||||
}
|
||||
|
||||
// LoadStageCompaction 按 stageKey 读取压缩摘要和水位线。
|
||||
func (a *AgentDAO) LoadStageCompaction(ctx context.Context, userID int, chatID string, stageKey string) (summary string, watermark int, err error) {
|
||||
stageKey = normalizeCompactionStageKey(stageKey)
|
||||
|
||||
var chat model.AgentChat
|
||||
err = a.db.WithContext(ctx).
|
||||
Select("compaction_summary", "compaction_watermark").
|
||||
Where("user_id = ? AND chat_id = ?", userID, chatID).
|
||||
First(&chat).Error
|
||||
if err != nil {
|
||||
return "", 0, err
|
||||
}
|
||||
|
||||
stages := loadStageCompactionStages(chat.CompactionSummary, chat.CompactionWatermark)
|
||||
if record, ok := stages[stageKey]; ok {
|
||||
return record.Summary, record.Watermark, nil
|
||||
}
|
||||
|
||||
return "", 0, nil
|
||||
}
|
||||
|
||||
// SaveStageCompaction 按 stageKey 保存压缩摘要和水位线。
|
||||
//
|
||||
// 1. 先读取现有摘要，避免覆盖其他阶段已经写入的数据。
|
||||
// 2. 再更新当前阶段对应的分桶内容。
|
||||
// 3. 最后整体回写 JSON envelope，并保留 execute 阶段的 legacy watermark 兼容字段。
|
||||
func (a *AgentDAO) SaveStageCompaction(ctx context.Context, userID int, chatID string, stageKey string, summary string, watermark int) error {
|
||||
stageKey = normalizeCompactionStageKey(stageKey)
|
||||
|
||||
var chat model.AgentChat
|
||||
err := a.db.WithContext(ctx).
|
||||
Select("compaction_summary", "compaction_watermark").
|
||||
Where("user_id = ? AND chat_id = ?", userID, chatID).
|
||||
First(&chat).Error
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
|
||||
stages := loadStageCompactionStages(chat.CompactionSummary, chat.CompactionWatermark)
|
||||
stages[stageKey] = stageCompactionRecord{
|
||||
Summary: strings.TrimSpace(summary),
|
||||
Watermark: watermark,
|
||||
}
|
||||
|
||||
payload, err := marshalStageCompactionStages(stages)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
|
||||
legacyWatermark := watermark
|
||||
if executeRecord, ok := stages["execute"]; ok {
|
||||
legacyWatermark = executeRecord.Watermark
|
||||
}
|
||||
|
||||
return a.db.WithContext(ctx).
|
||||
Model(&model.AgentChat{}).
|
||||
Where("user_id = ? AND chat_id = ?", userID, chatID).
|
||||
Updates(map[string]any{
|
||||
"compaction_summary": payload,
|
||||
"compaction_watermark": legacyWatermark,
|
||||
}).Error
|
||||
}
|
||||
|
||||
@@ -175,17 +175,13 @@ func (r *AgentResumeRequest) IsConnectionRecoverResume() bool {
|
||||
}
|
||||
|
||||
type ChatHistoryPersistPayload struct {
|
||||
UserID int `json:"user_id"`
|
||||
ConversationID string `json:"conversation_id"`
|
||||
Role string `json:"role"`
|
||||
Message string `json:"message"`
|
||||
ReasoningContent string `json:"reasoning_content,omitempty"`
|
||||
ReasoningDurationSeconds int `json:"reasoning_duration_seconds,omitempty"`
|
||||
RetryGroupID *string `json:"retry_group_id,omitempty"`
|
||||
RetryIndex *int `json:"retry_index,omitempty"`
|
||||
RetryFromUserMessageID *int `json:"retry_from_user_message_id,omitempty"`
|
||||
RetryFromAssistantMessageID *int `json:"retry_from_assistant_message_id,omitempty"`
|
||||
TokensConsumed int `json:"tokens_consumed"`
|
||||
UserID int `json:"user_id"`
|
||||
ConversationID string `json:"conversation_id"`
|
||||
Role string `json:"role"`
|
||||
Message string `json:"message"`
|
||||
ReasoningContent string `json:"reasoning_content,omitempty"`
|
||||
ReasoningDurationSeconds int `json:"reasoning_duration_seconds,omitempty"`
|
||||
TokensConsumed int `json:"tokens_consumed"`
|
||||
}
|
||||
|
||||
type ChatTokenUsageAdjustPayload struct {
|
||||
@@ -231,9 +227,6 @@ type GetConversationHistoryItem struct {
|
||||
CreatedAt *time.Time `json:"created_at,omitempty"`
|
||||
ReasoningContent string `json:"reasoning_content,omitempty"`
|
||||
ReasoningDurationSeconds int `json:"reasoning_duration_seconds,omitempty"`
|
||||
RetryGroupID *string `json:"retry_group_id"`
|
||||
RetryIndex *int `json:"retry_index"`
|
||||
RetryTotal *int `json:"retry_total"`
|
||||
}
|
||||
|
||||
type SchedulePlanPreviewCache struct {
|
||||
|
||||
@@ -479,24 +479,17 @@ LLM 的一次性文本输出通过 `SplitPseudoStreamText` 拆分成多个 chunk
|
||||
|
||||
### 9.1 prompt 构造模式
|
||||
|
||||
所有阶段共享 `buildStageMessages()` 函数：
|
||||
所有阶段现在统一共享 `buildUnifiedStageMessages()` 函数：
|
||||
|
||||
```
|
||||
System Prompt（节点专属）
|
||||
│
|
||||
v
|
||||
Pinned Blocks（置顶上下文块，作为独立 system 消息注入）
|
||||
│
|
||||
v
|
||||
Tool Schemas（工具 schema，作为独立 system 消息注入）
|
||||
│
|
||||
v
|
||||
History（对话历史，Tool 消息降级为 User 消息以兼容 API）
|
||||
│
|
||||
v
|
||||
User Prompt（节点专属用户提示）
|
||||
msg0(system) = 全局 system prompt + 阶段 system prompt + 工具简表
|
||||
msg1(assistant) = 对话历史 + 归档摘要
|
||||
msg2(assistant) = 阶段工作区
|
||||
msg3(system) = 阶段状态 + 记忆 + 本轮指令
|
||||
```
|
||||
|
||||
统一构造由 `StageMessagesConfig` 驱动，具体阶段只负责填充各自的 `Msg2Content`、`Msg3StageState` 和 `UserInstruction`。
|
||||
|
||||
### 9.2 各阶段 prompt 要点
|
||||
|
||||
| 阶段 | 核心指令 | 关键约束 |
|
||||
|
||||
@@ -93,7 +93,6 @@ type CommonState struct {
|
||||
// TaskClasses 本次排课涉及的任务类约束元数据（含日期、策略、时段预算等），
|
||||
// 在 Service 层从 DB 加载并注入，供 Plan prompt 直接消费，避免 LLM 因信息不足而追问用户。
|
||||
TaskClasses []schedule.TaskClassMeta `json:"task_classes,omitempty"`
|
||||
|
||||
// NeedsRoughBuild 由 Plan 节点在 plan_done 时写入，标记 Confirm 后是否需要走粗排节点。
|
||||
// 粗排节点执行完毕后会将此字段重置为 false。
|
||||
NeedsRoughBuild bool `json:"needs_rough_build,omitempty"`
|
||||
|
||||
@@ -9,6 +9,7 @@ import (
|
||||
newagentstream "github.com/LoveLosita/smartflow/backend/newAgent/stream"
|
||||
newagenttools "github.com/LoveLosita/smartflow/backend/newAgent/tools"
|
||||
schedule "github.com/LoveLosita/smartflow/backend/newAgent/tools/schedule"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
)
|
||||
|
||||
// AgentGraphRequest 描述一次 agent graph 运行的请求级输入。
|
||||
@@ -52,6 +53,14 @@ type RoughBuildFunc func(ctx context.Context, userID int, taskClassIDs []int) ([
|
||||
// 2. deliver 结束时再做最终覆盖写，保障收口状态一致。
|
||||
type WriteSchedulePreviewFunc func(ctx context.Context, state *schedule.ScheduleState, userID int, conversationID string, taskClassIDs []int) error
|
||||
|
||||
// PersistVisibleMessageFunc 是 newAgent 主循环逐条持久化可见消息的回调签名。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 只处理真正对用户可见的 assistant speak，不处理工具结果或内部纠错提示；
|
||||
// 2. 由节点在 AppendHistory 之后主动调用，让上层同步把这条消息写入 Redis + MySQL；
|
||||
// 3. 执行方可以做无损降级（例如 Redis 写失败只记日志），但应返回 error 便于上层记录。
|
||||
type PersistVisibleMessageFunc func(ctx context.Context, state *CommonState, msg *schema.Message) error
|
||||
|
||||
// AgentGraphDeps 描述 graph/node 层运行时真正依赖的可插拔能力。
|
||||
//
|
||||
// 设计目的：
|
||||
@@ -81,6 +90,10 @@ type AgentGraphDeps struct {
|
||||
// channel 携带已渲染的文本内容（非原始 ItemDTO），节点直接写入 pinned block。
|
||||
MemoryFuture chan string // buffered(1)，携带 renderMemoryPinnedContentByMode 的输出
|
||||
MemoryConsumed bool // 保证 channel 只读一次，后续 Execute ReAct 循环跳过等待
|
||||
|
||||
// PersistVisibleMessage 按 Service 注入，newAgent 每个节点产出的可见 speak
|
||||
// 都会在 AppendHistory 之后立刻调用这个回调，把消息同步落到 Redis + MySQL。
|
||||
PersistVisibleMessage PersistVisibleMessageFunc
|
||||
}
|
||||
|
||||
// --- 记忆 pinned block 常量（供 agentsvc 和 node 层共享） ---
|
||||
|
||||
@@ -81,9 +81,16 @@ type SchedulePersistor interface {
|
||||
}
|
||||
|
||||
// CompactionStore 定义上下文压缩的持久化接口。
|
||||
// 由 Service 层实现（组合 DAO + Redis Cache），注入到 ExecuteNodeInput。
|
||||
// 由 Service 层实现（组合 DAO + Redis Cache），注入到各阶段 NodeInput。
|
||||
type CompactionStore interface {
|
||||
LoadCompaction(ctx context.Context, userID int, chatID string) (summary string, watermark int, err error)
|
||||
SaveCompaction(ctx context.Context, userID int, chatID string, summary string, watermark int) error
|
||||
SaveContextTokenStats(ctx context.Context, userID int, chatID string, statsJSON string) error
|
||||
|
||||
// LoadStageCompaction 按 stageKey 加载压缩摘要和水位线。
|
||||
// stageKey 区分不同节点（如 "execute"/"plan"/"chat"/"deliver"），
|
||||
// 使各节点可以独立维护各自的压缩状态，互不覆盖。
|
||||
LoadStageCompaction(ctx context.Context, userID int, chatID string, stageKey string) (summary string, watermark int, err error)
|
||||
// SaveStageCompaction 按 stageKey 保存压缩摘要和水位线。
|
||||
SaveStageCompaction(ctx context.Context, userID int, chatID string, stageKey string, summary string, watermark int) error
|
||||
}
|
||||
|
||||
@@ -12,12 +12,12 @@ import (
|
||||
"github.com/LoveLosita/smartflow/backend/newAgent/tools/schedule"
|
||||
)
|
||||
|
||||
// AgentNodes 是 newAgent 通用图的节点容器。
|
||||
// AgentNodes 负责把 graph 层的节点调用统一转成 node 层真正的执行入口。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 负责把 node 层真正实现的方法统一暴露给 graph 注册；
|
||||
// 2. 负责收口"graph 只编排、node 真执行"的结构约束；
|
||||
// 3. 负责在每个节点执行成功后统一做状态持久化（Save/Delete）。
|
||||
// 1. 这里只做参数转发、依赖注入和状态落盘，不承载业务决策。
|
||||
// 2. 各节点真正的执行逻辑仍在对应的 RunXXXNode 内。
|
||||
// 3. 节点成功后统一保存快照，方便断线恢复。
|
||||
type AgentNodes struct{}
|
||||
|
||||
// NewAgentNodes 创建通用节点容器。
|
||||
@@ -25,104 +25,71 @@ func NewAgentNodes() *AgentNodes {
|
||||
return &AgentNodes{}
|
||||
}
|
||||
|
||||
// Chat 是聊天入口的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只做 graph -> node 的参数转接；
|
||||
// 2. 真正的入口逻辑仍由 RunChatNode 负责；
|
||||
// 3. Chat 的 Save 交给 Service 层处理，这里不做持久化。
|
||||
// Chat 负责把 graph 的 chat 节点请求转给 RunChatNode。
|
||||
func (n *AgentNodes) Chat(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("chat node: state is nil")
|
||||
}
|
||||
|
||||
// 注入工具 schema 到 ConversationContext，让路由决策更智能。
|
||||
if st.Deps.ToolRegistry != nil {
|
||||
schemas := st.Deps.ToolRegistry.Schemas()
|
||||
toolSchemas := make([]newagentmodel.ToolSchemaContext, len(schemas))
|
||||
for i, s := range schemas {
|
||||
toolSchemas[i] = newagentmodel.ToolSchemaContext{
|
||||
Name: s.Name,
|
||||
Desc: s.Desc,
|
||||
SchemaText: s.SchemaText,
|
||||
}
|
||||
}
|
||||
st.EnsureConversationContext().SetToolSchemas(toolSchemas)
|
||||
}
|
||||
// 1. Chat 阶段只负责路由与纯对话，不需要看到工具目录，避免能力细节干扰判断。
|
||||
st.EnsureConversationContext().SetToolSchemas(nil)
|
||||
|
||||
if err := RunChatNode(
|
||||
ctx,
|
||||
ChatNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
UserInput: st.Request.UserInput,
|
||||
ConfirmAction: st.Request.ConfirmAction,
|
||||
Client: st.Deps.ResolveChatClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
},
|
||||
); err != nil {
|
||||
if err := RunChatNode(ctx, ChatNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
UserInput: st.Request.UserInput,
|
||||
ConfirmAction: st.Request.ConfirmAction,
|
||||
Client: st.Deps.ResolveChatClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
CompactionStore: st.Deps.CompactionStore,
|
||||
PersistVisibleMessage: st.Deps.PersistVisibleMessage,
|
||||
}); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
saveAgentState(ctx, st)
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// Confirm 是确认阶段的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只做 graph -> node 的参数转接；
|
||||
// 2. 真正的确认逻辑仍由 RunConfirmNode 负责；
|
||||
// 3. 不需要 LLM Client — 确认内容由已有状态机械格式化。
|
||||
// 4. Confirm 执行成功后保存状态，因为它创建了 PendingInteraction。
|
||||
// Confirm 负责把 graph 的 confirm 节点请求转给 RunConfirmNode。
|
||||
func (n *AgentNodes) Confirm(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("confirm node: state is nil")
|
||||
}
|
||||
|
||||
if err := RunConfirmNode(
|
||||
ctx,
|
||||
ConfirmNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
},
|
||||
); err != nil {
|
||||
if err := RunConfirmNode(ctx, ConfirmNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
}); err != nil {
|
||||
return nil, err
|
||||
} else if st.Deps.WriteSchedulePreview != nil && st.ScheduleState == nil {
|
||||
flowState := st.EnsureFlowState()
|
||||
log.Printf("[WARN] deliver: schedule state is nil, skip preview write chat=%s", flowState.ConversationID)
|
||||
}
|
||||
|
||||
saveAgentState(ctx, st)
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// Plan 是规划阶段的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只做 graph -> node 的参数转接；
|
||||
// 2. 真正的单轮规划逻辑仍由 RunPlanNode 负责；
|
||||
// 3. Plan 执行成功后保存状态，支持意外断线恢复。
|
||||
// Plan 负责把 graph 的 plan 节点请求转给 RunPlanNode。
|
||||
func (n *AgentNodes) Plan(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("plan node: state is nil")
|
||||
}
|
||||
|
||||
// 等待后台记忆检索完成，注入最新记忆后再启动 Plan。
|
||||
// 等待后端记忆检索完成，再把最新结果注入上下文。
|
||||
ensureFreshMemory(st)
|
||||
|
||||
if err := RunPlanNode(
|
||||
ctx,
|
||||
PlanNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
UserInput: st.Request.UserInput,
|
||||
Client: st.Deps.ResolvePlanClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
ResumeNode: "plan",
|
||||
AlwaysExecute: st.Request.AlwaysExecute,
|
||||
ThinkingEnabled: st.Deps.ThinkingPlan,
|
||||
},
|
||||
); err != nil {
|
||||
if err := RunPlanNode(ctx, PlanNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
UserInput: st.Request.UserInput,
|
||||
Client: st.Deps.ResolvePlanClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
ResumeNode: "plan",
|
||||
AlwaysExecute: st.Request.AlwaysExecute,
|
||||
ThinkingEnabled: st.Deps.ThinkingPlan,
|
||||
CompactionStore: st.Deps.CompactionStore,
|
||||
PersistVisibleMessage: st.Deps.PersistVisibleMessage,
|
||||
}); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
@@ -130,12 +97,7 @@ func (n *AgentNodes) Plan(ctx context.Context, st *newagentmodel.AgentGraphState
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// RoughBuild 是粗排阶段的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 调用注入的 RoughBuildFunc 执行粗排算法；
|
||||
// 2. 把粗排结果写入 ScheduleState；
|
||||
// 3. 完成后保存状态，支持意外断线恢复。
|
||||
// RoughBuild 负责把 graph 的 rough_build 节点请求转给 RunRoughBuildNode。
|
||||
func (n *AgentNodes) RoughBuild(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("rough_build node: state is nil")
|
||||
@@ -149,48 +111,31 @@ func (n *AgentNodes) RoughBuild(ctx context.Context, st *newagentmodel.AgentGrap
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// Interrupt 是中断阶段的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只做 graph -> node 的参数转接；
|
||||
// 2. 真正的中断逻辑仍由 RunInterruptNode 负责；
|
||||
// 3. 不需要 LLM Client — 所有文本已在 PendingInteraction 里。
|
||||
// 4. 不需要 Save — 上游节点（Plan/Execute/Confirm）已经存过了。
|
||||
// Interrupt 负责把 graph 的 interrupt 节点请求转给 RunInterruptNode。
|
||||
func (n *AgentNodes) Interrupt(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("interrupt node: state is nil")
|
||||
}
|
||||
|
||||
if err := RunInterruptNode(
|
||||
ctx,
|
||||
InterruptNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
},
|
||||
); err != nil {
|
||||
if err := RunInterruptNode(ctx, InterruptNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
PersistVisibleMessage: st.Deps.PersistVisibleMessage,
|
||||
}); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// Execute 是执行阶段的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只做 graph -> node 的参数转接；
|
||||
// 2. 真正的单轮执行逻辑仍由 RunExecuteNode 负责。
|
||||
//
|
||||
// 设计原则：
|
||||
// 1. LLM 主导：LLM 自己判断 done_when 是否满足，自己决定何时推进/完成；
|
||||
// 2. 后端兜底：只做资源控制、安全兜底、证据记录；
|
||||
// 3. 不做硬校验：后端不质疑 LLM 的 advance/complete 决策。
|
||||
// 4. Execute 每轮执行成功后保存状态，支持意外断线恢复。
|
||||
// Execute 负责把 graph 的 execute 节点请求转给 RunExecuteNode。
|
||||
func (n *AgentNodes) Execute(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("execute node: state is nil")
|
||||
}
|
||||
|
||||
// 按需加载 ScheduleState（首次执行时从 DB 加载，后续复用内存中的 state）。
|
||||
// 1. 首次进入时按需加载日程状态，后续轮次复用内存状态。
|
||||
var scheduleState *schedule.ScheduleState
|
||||
if ss, loadErr := st.EnsureScheduleState(ctx); loadErr != nil {
|
||||
return nil, fmt.Errorf("execute node: 加载日程状态失败: %w", loadErr)
|
||||
@@ -198,7 +143,7 @@ func (n *AgentNodes) Execute(ctx context.Context, st *newagentmodel.AgentGraphSt
|
||||
scheduleState = ss
|
||||
}
|
||||
|
||||
// 注入工具 schema 到 ConversationContext，让 LLM 能看到可用工具列表。
|
||||
// 2. 把工具 schema 注入上下文，供 LLM 看到真实工具边界。
|
||||
if st.Deps.ToolRegistry != nil {
|
||||
schemas := st.Deps.ToolRegistry.Schemas()
|
||||
toolSchemas := make([]newagentmodel.ToolSchemaContext, len(schemas))
|
||||
@@ -212,28 +157,26 @@ func (n *AgentNodes) Execute(ctx context.Context, st *newagentmodel.AgentGraphSt
|
||||
st.EnsureConversationContext().SetToolSchemas(toolSchemas)
|
||||
}
|
||||
|
||||
// 等待后台记忆检索完成，注入最新记忆后再启动 Execute。
|
||||
// 3. 等待后端记忆检索结果，再把最新结果注入上下文。
|
||||
ensureFreshMemory(st)
|
||||
|
||||
if err := RunExecuteNode(
|
||||
ctx,
|
||||
ExecuteNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
UserInput: st.Request.UserInput,
|
||||
Client: st.Deps.ResolveExecuteClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
ResumeNode: "execute",
|
||||
ToolRegistry: st.Deps.ToolRegistry,
|
||||
ScheduleState: scheduleState,
|
||||
SchedulePersistor: st.Deps.SchedulePersistor,
|
||||
CompactionStore: st.Deps.CompactionStore,
|
||||
WriteSchedulePreview: st.Deps.WriteSchedulePreview,
|
||||
OriginalScheduleState: st.OriginalScheduleState,
|
||||
AlwaysExecute: st.Request.AlwaysExecute,
|
||||
ThinkingEnabled: st.Deps.ThinkingExecute,
|
||||
},
|
||||
); err != nil {
|
||||
if err := RunExecuteNode(ctx, ExecuteNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
UserInput: st.Request.UserInput,
|
||||
Client: st.Deps.ResolveExecuteClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
ResumeNode: "execute",
|
||||
ToolRegistry: st.Deps.ToolRegistry,
|
||||
ScheduleState: scheduleState,
|
||||
SchedulePersistor: st.Deps.SchedulePersistor,
|
||||
CompactionStore: st.Deps.CompactionStore,
|
||||
WriteSchedulePreview: st.Deps.WriteSchedulePreview,
|
||||
OriginalScheduleState: st.OriginalScheduleState,
|
||||
AlwaysExecute: st.Request.AlwaysExecute,
|
||||
ThinkingEnabled: st.Deps.ThinkingExecute,
|
||||
PersistVisibleMessage: st.Deps.PersistVisibleMessage,
|
||||
}); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
@@ -241,12 +184,7 @@ func (n *AgentNodes) Execute(ctx context.Context, st *newagentmodel.AgentGraphSt
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// OrderGuard 是顺序守卫阶段的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 只负责调用 RunOrderGuardNode 做 suggested 相对顺序校验；
|
||||
// 2. 不负责交付文案生成，校验结果统一交给 Deliver 节点收口；
|
||||
// 3. 节点执行后保存状态，保证异常中断后仍可复盘守卫结果。
|
||||
// OrderGuard 负责把 graph 的 order_guard 节点请求转给 RunOrderGuardNode。
|
||||
func (n *AgentNodes) OrderGuard(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("order_guard node: state is nil")
|
||||
@@ -260,38 +198,32 @@ func (n *AgentNodes) OrderGuard(ctx context.Context, st *newagentmodel.AgentGrap
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// Deliver 是交付阶段的正式节点方法。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只做 graph -> node 的参数转接；
|
||||
// 2. 真正的交付逻辑仍由 RunDeliverNode 负责；
|
||||
// 3. 调 LLM 生成任务总结，失败时降级到机械格式化。
|
||||
// 4. 任务完成后保存最终状态到 Redis（2h TTL），支持断线恢复和 MySQL outbox 异步持久化。
|
||||
// Deliver 负责把 graph 的 deliver 节点请求转给 RunDeliverNode。
|
||||
func (n *AgentNodes) Deliver(ctx context.Context, st *newagentmodel.AgentGraphState) (*newagentmodel.AgentGraphState, error) {
|
||||
if st == nil {
|
||||
return nil, errors.New("deliver node: state is nil")
|
||||
}
|
||||
|
||||
if err := RunDeliverNode(
|
||||
ctx,
|
||||
DeliverNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
Client: st.Deps.ResolveDeliverClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
ThinkingEnabled: st.Deps.ThinkingDeliver,
|
||||
},
|
||||
); err != nil {
|
||||
// 1. Deliver 只做最终收口总结，不需要工具目录，避免无关能力信息污染总结。
|
||||
st.EnsureConversationContext().SetToolSchemas(nil)
|
||||
|
||||
if err := RunDeliverNode(ctx, DeliverNodeInput{
|
||||
RuntimeState: st.EnsureRuntimeState(),
|
||||
ConversationContext: st.EnsureConversationContext(),
|
||||
Client: st.Deps.ResolveDeliverClient(),
|
||||
ChunkEmitter: st.EnsureChunkEmitter(),
|
||||
ThinkingEnabled: st.Deps.ThinkingDeliver,
|
||||
CompactionStore: st.Deps.CompactionStore,
|
||||
PersistVisibleMessage: st.Deps.PersistVisibleMessage,
|
||||
}); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
// 任务完成后写排程预览缓存：只有走到 Deliver 才代表排程结果已稳定，
|
||||
// 中断（confirm/ask_user）路径不写，避免把中间态暴露给前端。
|
||||
// 只有真正完成时才写入排程预览，避免中间态污染前端展示。
|
||||
if st.Deps.WriteSchedulePreview != nil && st.ScheduleState != nil {
|
||||
flowState := st.EnsureFlowState()
|
||||
if flowState != nil && flowState.IsCompleted() {
|
||||
if err := st.Deps.WriteSchedulePreview(ctx, st.ScheduleState, flowState.UserID, flowState.ConversationID, flowState.TaskClassIDs); err != nil {
|
||||
// 写缓存失败不阻断主流程，降级为仅 log。
|
||||
log.Printf("[WARN] deliver: 写入排程预览缓存失败 chat=%s: %v", flowState.ConversationID, err)
|
||||
}
|
||||
} else if flowState != nil {
|
||||
@@ -303,19 +235,16 @@ func (n *AgentNodes) Deliver(ctx context.Context, st *newagentmodel.AgentGraphSt
|
||||
return st, nil
|
||||
}
|
||||
|
||||
// --- 记忆预取消费辅助 ---
|
||||
|
||||
// ensureFreshMemory 等待后台记忆检索完成，将最新结果注入 ConversationContext。
|
||||
// ensureFreshMemory 等待后端记忆检索完成，并把最新结果写入 ConversationContext。
|
||||
//
|
||||
// 设计说明：
|
||||
// 1. 只在首次调用时等待 channel（最多 500ms），后续调用直接跳过；
|
||||
// 2. 覆盖 ConversationContext 中已有的缓存记忆（UpsertPinnedBlock 按 key 覆盖）；
|
||||
// 3. timeout 后保留缓存记忆不替换，保证 Execute ReAct 循环不会因超时丢失记忆。
|
||||
// 1. 只在首次调用时等待 channel，后续调用直接跳过。
|
||||
// 2. 超时后保留原有上下文，不额外覆盖。
|
||||
// 3. 记忆为空时也不做额外写入，避免污染 prompt。
|
||||
func ensureFreshMemory(st *newagentmodel.AgentGraphState) {
|
||||
if st == nil || st.Deps.MemoryConsumed || st.Deps.MemoryFuture == nil {
|
||||
return
|
||||
}
|
||||
st.Deps.MemoryConsumed = true // 标记已消费，后续调用直接跳过
|
||||
st.Deps.MemoryConsumed = true
|
||||
|
||||
select {
|
||||
case content := <-st.Deps.MemoryFuture:
|
||||
@@ -327,20 +256,11 @@ func ensureFreshMemory(st *newagentmodel.AgentGraphState) {
|
||||
})
|
||||
}
|
||||
case <-time.After(newagentmodel.MemoryFreshTimeout):
|
||||
// timeout：保留 ConversationContext 中已有的缓存记忆，不做额外操作
|
||||
// 超时后保留原有上下文即可。
|
||||
}
|
||||
}
|
||||
|
||||
// --- 持久化辅助 ---
|
||||
|
||||
// saveAgentState 在节点执行成功后，将当前运行态快照保存到 Redis。
|
||||
//
|
||||
// 设计原则：
|
||||
// 1. Save 失败只记日志，不中断 Graph 流程；
|
||||
// 2. StateStore 为空时静默跳过（骨架期 / 测试环境）；
|
||||
// 3. conversationID 为空时也静默跳过，避免写入无效 key。
|
||||
//
|
||||
// TODO: 接入项目统一的日志框架后，把 _ = err 改成结构化日志。
|
||||
// saveAgentState 在节点成功执行后保存运行快照。
|
||||
func saveAgentState(ctx context.Context, st *newagentmodel.AgentGraphState) {
|
||||
if st == nil {
|
||||
return
|
||||
@@ -370,14 +290,7 @@ func saveAgentState(ctx context.Context, st *newagentmodel.AgentGraphState) {
|
||||
_ = store.Save(ctx, flowState.ConversationID, snapshot)
|
||||
}
|
||||
|
||||
// deleteAgentState 在任务完成后，删除 Redis 中的运行态快照。
|
||||
//
|
||||
// 设计原则：
|
||||
// 1. Delete 失败只记日志，不中断 Graph 流程；
|
||||
// 2. 删除是幂等的，key 不存在也视为成功；
|
||||
// 3. StateStore 为空时静默跳过。
|
||||
//
|
||||
// TODO: 接入项目统一的日志框架后，把 _ = err 改成结构化日志。
|
||||
// deleteAgentState 在任务完成后删除运行快照。
|
||||
func deleteAgentState(ctx context.Context, st *newagentmodel.AgentGraphState) {
|
||||
if st == nil {
|
||||
return
|
||||
|
||||
@@ -45,12 +45,14 @@ const (
|
||||
// 3. ConversationContext 提供历史对话；
|
||||
// 4. ConfirmAction 仅在 confirm 恢复场景下由前端传入 "accept" / "reject"。
|
||||
type ChatNodeInput struct {
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
UserInput string
|
||||
ConfirmAction string
|
||||
Client *infrallm.Client
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
UserInput string
|
||||
ConfirmAction string
|
||||
Client *infrallm.Client
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
CompactionStore newagentmodel.CompactionStore // 上下文压缩持久化
|
||||
PersistVisibleMessage newagentmodel.PersistVisibleMessageFunc
|
||||
}
|
||||
|
||||
// RunChatNode 执行一轮聊天节点逻辑。
|
||||
@@ -94,6 +96,15 @@ func RunChatNode(ctx context.Context, input ChatNodeInput) error {
|
||||
}
|
||||
nonce := uuid.NewString()
|
||||
messages := newagentprompt.BuildChatRoutingMessages(conversationContext, input.UserInput, flowState, nonce)
|
||||
messages = compactUnifiedMessagesIfNeeded(ctx, messages, UnifiedCompactInput{
|
||||
Client: input.Client,
|
||||
CompactionStore: input.CompactionStore,
|
||||
FlowState: flowState,
|
||||
Emitter: emitter,
|
||||
StageName: chatStageName,
|
||||
StatusBlockID: chatStatusBlockID,
|
||||
})
|
||||
logNodeLLMContext(chatStageName, "routing", flowState, messages)
|
||||
|
||||
reader, err := input.Client.Stream(ctx, messages, infrallm.GenerateOptions{
|
||||
Temperature: 0.7,
|
||||
@@ -281,7 +292,7 @@ func handleDirectReplyStream(
|
||||
if effectiveThinking {
|
||||
return handleThinkingReplyStream(ctx, reader, input, emitter, conversationContext, flowState)
|
||||
}
|
||||
return handleDirectReplyContinueStream(ctx, reader, emitter, conversationContext, flowState, firstVisible)
|
||||
return handleDirectReplyContinueStream(ctx, reader, input, emitter, conversationContext, flowState, firstVisible)
|
||||
}
|
||||
|
||||
// handleThinkingReplyStream 处理需要思考的回复：关闭路由流 → 第二次 thinking 流式调用。
|
||||
@@ -295,7 +306,16 @@ func handleThinkingReplyStream(
|
||||
) error {
|
||||
_ = reader.Close()
|
||||
|
||||
deepMessages := newagentprompt.BuildDeepAnswerMessages(conversationContext, input.UserInput)
|
||||
deepMessages := newagentprompt.BuildDeepAnswerMessages(flowState, conversationContext, input.UserInput)
|
||||
deepMessages = compactUnifiedMessagesIfNeeded(ctx, deepMessages, UnifiedCompactInput{
|
||||
Client: input.Client,
|
||||
CompactionStore: input.CompactionStore,
|
||||
FlowState: flowState,
|
||||
Emitter: emitter,
|
||||
StageName: chatStageName,
|
||||
StatusBlockID: chatStatusBlockID,
|
||||
})
|
||||
logNodeLLMContext(chatStageName, "direct_reply_thinking", flowState, deepMessages)
|
||||
deepReader, err := input.Client.Stream(ctx, deepMessages, infrallm.GenerateOptions{
|
||||
Temperature: 0.5,
|
||||
MaxTokens: 2000,
|
||||
@@ -322,6 +342,7 @@ func handleThinkingReplyStream(
|
||||
deepText = strings.TrimSpace(deepText)
|
||||
if deepText != "" {
|
||||
conversationContext.AppendHistory(schema.AssistantMessage(deepText, nil))
|
||||
persistVisibleAssistantMessage(ctx, input.PersistVisibleMessage, flowState, schema.AssistantMessage(deepText, nil))
|
||||
}
|
||||
|
||||
flowState.Phase = newagentmodel.PhaseChatting
|
||||
@@ -332,6 +353,7 @@ func handleThinkingReplyStream(
|
||||
func handleDirectReplyContinueStream(
|
||||
ctx context.Context,
|
||||
reader infrallm.StreamReader,
|
||||
input ChatNodeInput,
|
||||
emitter *newagentstream.ChunkEmitter,
|
||||
conversationContext *newagentmodel.ConversationContext,
|
||||
flowState *newagentmodel.CommonState,
|
||||
@@ -370,7 +392,9 @@ func handleDirectReplyContinueStream(
|
||||
|
||||
text := fullText.String()
|
||||
if strings.TrimSpace(text) != "" {
|
||||
conversationContext.AppendHistory(schema.AssistantMessage(text, nil))
|
||||
msg := schema.AssistantMessage(text, nil)
|
||||
conversationContext.AppendHistory(msg)
|
||||
persistVisibleAssistantMessage(ctx, input.PersistVisibleMessage, flowState, msg)
|
||||
}
|
||||
|
||||
flowState.Phase = newagentmodel.PhaseChatting
|
||||
@@ -568,7 +592,16 @@ func handleDeepAnswerStream(
|
||||
if effectiveThinking {
|
||||
thinkingOpt = infrallm.ThinkingModeEnabled
|
||||
}
|
||||
deepMessages := newagentprompt.BuildDeepAnswerMessages(conversationContext, input.UserInput)
|
||||
deepMessages := newagentprompt.BuildDeepAnswerMessages(flowState, conversationContext, input.UserInput)
|
||||
deepMessages = compactUnifiedMessagesIfNeeded(ctx, deepMessages, UnifiedCompactInput{
|
||||
Client: input.Client,
|
||||
CompactionStore: input.CompactionStore,
|
||||
FlowState: flowState,
|
||||
Emitter: emitter,
|
||||
StageName: chatStageName,
|
||||
StatusBlockID: chatStatusBlockID,
|
||||
})
|
||||
logNodeLLMContext(chatStageName, "deep_answer", flowState, deepMessages)
|
||||
deepReader, err := input.Client.Stream(ctx, deepMessages, infrallm.GenerateOptions{
|
||||
Temperature: 0.5,
|
||||
MaxTokens: 2000,
|
||||
@@ -601,7 +634,9 @@ func handleDeepAnswerStream(
|
||||
}
|
||||
|
||||
// 4. 完整回复写入 history。
|
||||
conversationContext.AppendHistory(schema.AssistantMessage(deepText, nil))
|
||||
msg := schema.AssistantMessage(deepText, nil)
|
||||
conversationContext.AppendHistory(msg)
|
||||
persistVisibleAssistantMessage(ctx, input.PersistVisibleMessage, flowState, msg)
|
||||
|
||||
flowState.Phase = newagentmodel.PhaseChatting
|
||||
return nil
|
||||
|
||||
@@ -28,11 +28,13 @@ const (
|
||||
// 3. ConversationContext 提供执行阶段的对话历史；
|
||||
// 4. 交付完成后标记流程结束。
|
||||
type DeliverNodeInput struct {
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
Client *infrallm.Client
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
ThinkingEnabled bool // 是否开启 thinking，由 config.yaml 的 agent.thinking.deliver 注入
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
Client *infrallm.Client
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
ThinkingEnabled bool // 是否开启 thinking，由 config.yaml 的 agent.thinking.deliver 注入
|
||||
CompactionStore newagentmodel.CompactionStore // 上下文压缩持久化
|
||||
PersistVisibleMessage newagentmodel.PersistVisibleMessageFunc
|
||||
}
|
||||
|
||||
// RunDeliverNode 执行一轮交付节点逻辑。
|
||||
@@ -65,10 +67,11 @@ func RunDeliverNode(ctx context.Context, input DeliverNodeInput) error {
|
||||
}
|
||||
|
||||
// 2. 调 LLM 生成交付总结。
|
||||
summary := generateDeliverSummary(ctx, input.Client, flowState, conversationContext, input.ThinkingEnabled)
|
||||
summary := generateDeliverSummary(ctx, input.Client, flowState, conversationContext, input.ThinkingEnabled, input.CompactionStore, emitter)
|
||||
|
||||
// 3. 伪流式推送总结。
|
||||
if strings.TrimSpace(summary) != "" {
|
||||
msg := schema.AssistantMessage(summary, nil)
|
||||
if err := emitter.EmitPseudoAssistantText(
|
||||
ctx,
|
||||
deliverSpeakBlockID,
|
||||
@@ -78,7 +81,8 @@ func RunDeliverNode(ctx context.Context, input DeliverNodeInput) error {
|
||||
); err != nil {
|
||||
return fmt.Errorf("交付总结推送失败: %w", err)
|
||||
}
|
||||
conversationContext.AppendHistory(schema.AssistantMessage(summary, nil))
|
||||
conversationContext.AppendHistory(msg)
|
||||
persistVisibleAssistantMessage(ctx, input.PersistVisibleMessage, flowState, msg)
|
||||
}
|
||||
|
||||
// 4. 推送最终完成状态。
|
||||
@@ -100,6 +104,8 @@ func generateDeliverSummary(
|
||||
flowState *newagentmodel.CommonState,
|
||||
conversationContext *newagentmodel.ConversationContext,
|
||||
thinkingEnabled bool,
|
||||
compactionStore newagentmodel.CompactionStore,
|
||||
emitter *newagentstream.ChunkEmitter,
|
||||
) string {
|
||||
if flowState != nil {
|
||||
switch {
|
||||
@@ -115,6 +121,15 @@ func generateDeliverSummary(
|
||||
}
|
||||
|
||||
messages := newagentprompt.BuildDeliverMessages(flowState, conversationContext)
|
||||
messages = compactUnifiedMessagesIfNeeded(ctx, messages, UnifiedCompactInput{
|
||||
Client: client,
|
||||
CompactionStore: compactionStore,
|
||||
FlowState: flowState,
|
||||
Emitter: emitter,
|
||||
StageName: deliverStageName,
|
||||
StatusBlockID: deliverStatusBlockID,
|
||||
})
|
||||
logNodeLLMContext(deliverStageName, "summarizing", flowState, messages)
|
||||
result, err := client.GenerateText(
|
||||
ctx,
|
||||
messages,
|
||||
|
||||
@@ -60,6 +60,7 @@ type ExecuteNodeInput struct {
|
||||
OriginalScheduleState *schedule.ScheduleState
|
||||
AlwaysExecute bool // true 时写工具跳过确认闸门直接执行
|
||||
ThinkingEnabled bool // 是否开启 thinking，由 config.yaml 的 agent.thinking.execute 注入
|
||||
PersistVisibleMessage newagentmodel.PersistVisibleMessageFunc
|
||||
}
|
||||
|
||||
// ExecuteRoundObservation 记录执行阶段每轮的关键观察。
|
||||
@@ -184,19 +185,16 @@ func RunExecuteNode(ctx context.Context, input ExecuteNodeInput) error {
|
||||
messages := newagentprompt.BuildExecuteMessages(flowState, conversationContext)
|
||||
|
||||
// 5.1 Token 预算检查 & 上下文压缩。
|
||||
messages = compactExecuteMessagesIfNeeded(
|
||||
ctx, messages, input, flowState, emitter,
|
||||
)
|
||||
messages = compactUnifiedMessagesIfNeeded(ctx, messages, UnifiedCompactInput{
|
||||
Client: input.Client,
|
||||
CompactionStore: input.CompactionStore,
|
||||
FlowState: flowState,
|
||||
Emitter: emitter,
|
||||
StageName: executeStageName,
|
||||
StatusBlockID: executeStatusBlockID,
|
||||
})
|
||||
|
||||
log.Printf(
|
||||
"[DEBUG] execute LLM context begin chat=%s round=%d message_count=%d\n%s\n[DEBUG] execute LLM context end chat=%s round=%d",
|
||||
flowState.ConversationID,
|
||||
flowState.RoundUsed,
|
||||
len(messages),
|
||||
formatExecuteLLMMessagesForDebug(messages),
|
||||
flowState.ConversationID,
|
||||
flowState.RoundUsed,
|
||||
)
|
||||
logNodeLLMContext(executeStageName, "decision", flowState, messages)
|
||||
decision, rawResult, err := infrallm.GenerateJSON[newagentmodel.ExecuteDecision](
|
||||
ctx,
|
||||
input.Client,
|
||||
@@ -337,6 +335,7 @@ func RunExecuteNode(ctx context.Context, input ExecuteNodeInput) error {
|
||||
|
||||
if !isConfirmWithCard && !isAskUser && !isAbort {
|
||||
// 推流给前端
|
||||
msg := schema.AssistantMessage(speakText, nil)
|
||||
if err := emitter.EmitPseudoAssistantText(
|
||||
ctx,
|
||||
executeSpeakBlockID,
|
||||
@@ -346,6 +345,7 @@ func RunExecuteNode(ctx context.Context, input ExecuteNodeInput) error {
|
||||
); err != nil {
|
||||
return fmt.Errorf("执行文案推送失败: %w", err)
|
||||
}
|
||||
persistVisibleAssistantMessage(ctx, input.PersistVisibleMessage, flowState, msg)
|
||||
}
|
||||
// 1. confirm / ask_user 的 speak 仍要写入历史，避免下一轮 LLM 丢失自己的执行上下文。
|
||||
// 2. abort 不在这里写历史，避免先输出中间 speak，再在 deliver 收到第二份终止文案。
|
||||
@@ -1674,79 +1674,3 @@ func flattenForLog(text string) string {
|
||||
text = strings.ReplaceAll(text, "\r", " ")
|
||||
return strings.TrimSpace(text)
|
||||
}
|
||||
|
||||
// formatExecuteLLMMessagesForDebug 将本轮送入 LLM 的完整消息上下文展开成可读多行日志。
|
||||
//
|
||||
// 说明：
|
||||
// 1. 按消息索引逐条输出，便于和上游上下文构造步骤逐项对齐；
|
||||
// 2. 完整输出 content / reasoning_content / tool_calls / extra，不做截断；
|
||||
// 3. 仅用于调试打点，不参与业务决策。
|
||||
func formatExecuteLLMMessagesForDebug(messages []*schema.Message) string {
|
||||
if len(messages) == 0 {
|
||||
return "(empty messages)"
|
||||
}
|
||||
|
||||
var sb strings.Builder
|
||||
for i, msg := range messages {
|
||||
sb.WriteString(fmt.Sprintf("----- message[%d] -----\n", i))
|
||||
if msg == nil {
|
||||
sb.WriteString("role: <nil>\n\n")
|
||||
continue
|
||||
}
|
||||
|
||||
sb.WriteString(fmt.Sprintf("role: %s\n", msg.Role))
|
||||
|
||||
if strings.TrimSpace(msg.ToolCallID) != "" {
|
||||
sb.WriteString(fmt.Sprintf("tool_call_id: %s\n", msg.ToolCallID))
|
||||
}
|
||||
if strings.TrimSpace(msg.ToolName) != "" {
|
||||
sb.WriteString(fmt.Sprintf("tool_name: %s\n", msg.ToolName))
|
||||
}
|
||||
|
||||
if len(msg.ToolCalls) > 0 {
|
||||
sb.WriteString("tool_calls:\n")
|
||||
for j, call := range msg.ToolCalls {
|
||||
sb.WriteString(fmt.Sprintf(" - [%d] id=%s type=%s function=%s\n", j, call.ID, call.Type, call.Function.Name))
|
||||
sb.WriteString(" arguments:\n")
|
||||
sb.WriteString(indentMultilineForDebug(call.Function.Arguments, " "))
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
}
|
||||
|
||||
if strings.TrimSpace(msg.ReasoningContent) != "" {
|
||||
sb.WriteString("reasoning_content:\n")
|
||||
sb.WriteString(indentMultilineForDebug(msg.ReasoningContent, " "))
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
|
||||
sb.WriteString("content:\n")
|
||||
sb.WriteString(indentMultilineForDebug(msg.Content, " "))
|
||||
sb.WriteString("\n")
|
||||
|
||||
if len(msg.Extra) > 0 {
|
||||
sb.WriteString("extra:\n")
|
||||
raw, err := json.MarshalIndent(msg.Extra, "", " ")
|
||||
if err != nil {
|
||||
sb.WriteString(indentMultilineForDebug("<marshal_error>", " "))
|
||||
} else {
|
||||
sb.WriteString(indentMultilineForDebug(string(raw), " "))
|
||||
}
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
return sb.String()
|
||||
}
|
||||
|
||||
// indentMultilineForDebug 为多行文本统一添加前缀缩进，避免日志折行后难以阅读。
|
||||
func indentMultilineForDebug(text, prefix string) string {
|
||||
if text == "" {
|
||||
return prefix + "<empty>"
|
||||
}
|
||||
lines := strings.Split(text, "\n")
|
||||
for i := range lines {
|
||||
lines[i] = prefix + lines[i]
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
@@ -1,197 +0,0 @@
|
||||
package newagentnode
|
||||
|
||||
import (
|
||||
"context"
|
||||
"encoding/json"
|
||||
"fmt"
|
||||
"log"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
newagentprompt "github.com/LoveLosita/smartflow/backend/newAgent/prompt"
|
||||
newagentstream "github.com/LoveLosita/smartflow/backend/newAgent/stream"
|
||||
"github.com/LoveLosita/smartflow/backend/pkg"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
)
|
||||
|
||||
// compactExecuteMessagesIfNeeded 检查 Execute prompt 的 token 预算，
|
||||
// 超限时对 msg1（历史对话）和 msg2（ReAct Loop）执行 LLM 压缩。
|
||||
//
|
||||
// 消息布局约定（由 BuildExecuteMessages 返回）：
|
||||
//
|
||||
// [0] system — msg0: 系统规则
|
||||
// [1] assistant — msg1: 历史对话上下文
|
||||
// [2] assistant — msg2: 当轮 ReAct Loop 记录
|
||||
// [3] system — msg3: 当前状态 + 用户提示
|
||||
func compactExecuteMessagesIfNeeded(
|
||||
ctx context.Context,
|
||||
messages []*schema.Message,
|
||||
input ExecuteNodeInput,
|
||||
flowState *newagentmodel.CommonState,
|
||||
emitter *newagentstream.ChunkEmitter,
|
||||
) []*schema.Message {
|
||||
if len(messages) != 4 {
|
||||
return messages
|
||||
}
|
||||
|
||||
// 提取四条消息的文本内容
|
||||
msg0 := messages[0].Content
|
||||
msg1 := messages[1].Content
|
||||
msg2 := messages[2].Content
|
||||
msg3 := messages[3].Content
|
||||
|
||||
// Token 预算检查
|
||||
breakdown, overBudget, needCompactMsg1, needCompactMsg2 := pkg.CheckExecuteTokenBudget(msg0, msg1, msg2, msg3)
|
||||
|
||||
log.Printf(
|
||||
"[COMPACT] token budget check: total=%d budget=%d over=%v compactMsg1=%v compactMsg2=%v (msg0=%d msg1=%d msg2=%d msg3=%d)",
|
||||
breakdown.Total, breakdown.Budget, overBudget, needCompactMsg1, needCompactMsg2,
|
||||
breakdown.Msg0, breakdown.Msg1, breakdown.Msg2, breakdown.Msg3,
|
||||
)
|
||||
|
||||
if !overBudget {
|
||||
// 未超限，记录 token 分布后直接返回
|
||||
saveTokenStats(ctx, input, flowState, breakdown)
|
||||
return messages
|
||||
}
|
||||
|
||||
// ---- msg1 压缩 ----
|
||||
if needCompactMsg1 {
|
||||
msg1 = compactMsg1IfNeeded(ctx, input, flowState, emitter, msg1)
|
||||
messages[1].Content = msg1
|
||||
// 压缩 msg1 后重算预算
|
||||
breakdown = pkg.EstimateExecuteMessagesTokens(msg0, msg1, msg2, msg3)
|
||||
}
|
||||
|
||||
// ---- msg2 压缩 ----
|
||||
if needCompactMsg2 || breakdown.Total > pkg.ExecuteTokenBudget {
|
||||
msg2 = compactMsg2IfNeeded(ctx, input, flowState, emitter, msg2)
|
||||
messages[2].Content = msg2
|
||||
breakdown = pkg.EstimateExecuteMessagesTokens(msg0, msg1, msg2, msg3)
|
||||
}
|
||||
|
||||
// 记录最终 token 分布
|
||||
saveTokenStats(ctx, input, flowState, breakdown)
|
||||
|
||||
log.Printf(
|
||||
"[COMPACT] after compaction: total=%d budget=%d (msg0=%d msg1=%d msg2=%d msg3=%d)",
|
||||
breakdown.Total, breakdown.Budget,
|
||||
breakdown.Msg0, breakdown.Msg1, breakdown.Msg2, breakdown.Msg3,
|
||||
)
|
||||
return messages
|
||||
}
|
||||
|
||||
// compactMsg1IfNeeded 对 msg1（历史对话）执行 LLM 压缩。
|
||||
func compactMsg1IfNeeded(
|
||||
ctx context.Context,
|
||||
input ExecuteNodeInput,
|
||||
flowState *newagentmodel.CommonState,
|
||||
emitter *newagentstream.ChunkEmitter,
|
||||
msg1 string,
|
||||
) string {
|
||||
compactionStore := input.CompactionStore
|
||||
if compactionStore == nil {
|
||||
log.Printf("[COMPACT] CompactionStore is nil, skip msg1 compaction")
|
||||
return msg1
|
||||
}
|
||||
|
||||
// 加载已有压缩摘要
|
||||
existingSummary, _, err := compactionStore.LoadCompaction(ctx, flowState.UserID, flowState.ConversationID)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT] load existing compaction failed: %v, proceed without cache", err)
|
||||
}
|
||||
|
||||
// SSE: 压缩开始
|
||||
tokenBefore := pkg.EstimateTextTokens(msg1)
|
||||
_ = emitter.EmitStatus(
|
||||
executeStatusBlockID, "compact_msg1", "context_compact_start",
|
||||
fmt.Sprintf("正在压缩对话历史（%d tokens）...", tokenBefore),
|
||||
false,
|
||||
)
|
||||
|
||||
// 调用 LLM 压缩
|
||||
newSummary, err := newagentprompt.CompactMsg1(ctx, input.Client, msg1, existingSummary)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT] compact msg1 failed: %v", err)
|
||||
_ = emitter.EmitStatus(
|
||||
executeStatusBlockID, "compact_msg1", "context_compact_done",
|
||||
"对话历史压缩失败，使用原始文本",
|
||||
false,
|
||||
)
|
||||
return msg1
|
||||
}
|
||||
|
||||
// SSE: 压缩完成
|
||||
tokenAfter := pkg.EstimateTextTokens(newSummary)
|
||||
_ = emitter.EmitStatus(
|
||||
executeStatusBlockID, "compact_msg1", "context_compact_done",
|
||||
fmt.Sprintf("对话历史已压缩：%d → %d tokens", tokenBefore, tokenAfter),
|
||||
false,
|
||||
)
|
||||
|
||||
// 持久化压缩结果
|
||||
if err := compactionStore.SaveCompaction(ctx, flowState.UserID, flowState.ConversationID, newSummary, flowState.RoundUsed); err != nil {
|
||||
log.Printf("[COMPACT] save compaction failed: %v", err)
|
||||
}
|
||||
|
||||
return newSummary
|
||||
}
|
||||
|
||||
// compactMsg2IfNeeded 对 msg2（ReAct Loop 记录）执行 LLM 压缩。
|
||||
func compactMsg2IfNeeded(
|
||||
ctx context.Context,
|
||||
input ExecuteNodeInput,
|
||||
flowState *newagentmodel.CommonState,
|
||||
emitter *newagentstream.ChunkEmitter,
|
||||
msg2 string,
|
||||
) string {
|
||||
// SSE: 压缩开始
|
||||
tokenBefore := pkg.EstimateTextTokens(msg2)
|
||||
_ = emitter.EmitStatus(
|
||||
executeStatusBlockID, "compact_msg2", "context_compact_start",
|
||||
fmt.Sprintf("正在压缩执行记录（%d tokens）...", tokenBefore),
|
||||
false,
|
||||
)
|
||||
|
||||
// 调用 LLM 压缩
|
||||
compressed, err := newagentprompt.CompactMsg2(ctx, input.Client, msg2)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT] compact msg2 failed: %v", err)
|
||||
_ = emitter.EmitStatus(
|
||||
executeStatusBlockID, "compact_msg2", "context_compact_done",
|
||||
"执行记录压缩失败，使用原始文本",
|
||||
false,
|
||||
)
|
||||
return msg2
|
||||
}
|
||||
|
||||
// SSE: 压缩完成
|
||||
tokenAfter := pkg.EstimateTextTokens(compressed)
|
||||
_ = emitter.EmitStatus(
|
||||
executeStatusBlockID, "compact_msg2", "context_compact_done",
|
||||
fmt.Sprintf("执行记录已压缩：%d → %d tokens", tokenBefore, tokenAfter),
|
||||
false,
|
||||
)
|
||||
|
||||
return compressed
|
||||
}
|
||||
|
||||
// saveTokenStats 持久化当前 token 分布到 DB。
|
||||
func saveTokenStats(
|
||||
ctx context.Context,
|
||||
input ExecuteNodeInput,
|
||||
flowState *newagentmodel.CommonState,
|
||||
breakdown pkg.ExecuteTokenBreakdown,
|
||||
) {
|
||||
compactionStore := input.CompactionStore
|
||||
if compactionStore == nil {
|
||||
return
|
||||
}
|
||||
statsJSON, err := json.Marshal(breakdown)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT] marshal token stats failed: %v", err)
|
||||
return
|
||||
}
|
||||
if err := compactionStore.SaveContextTokenStats(ctx, flowState.UserID, flowState.ConversationID, string(statsJSON)); err != nil {
|
||||
log.Printf("[COMPACT] save token stats failed: %v", err)
|
||||
}
|
||||
}
|
||||
@@ -24,9 +24,10 @@ const (
|
||||
// 2. RuntimeState 提供 PendingInteraction；
|
||||
// 3. ChunkEmitter 负责推送收尾消息。
|
||||
type InterruptNodeInput struct {
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
PersistVisibleMessage newagentmodel.PersistVisibleMessageFunc
|
||||
}
|
||||
|
||||
// RunInterruptNode 执行一轮中断节点逻辑。
|
||||
@@ -55,7 +56,7 @@ func RunInterruptNode(ctx context.Context, input InterruptNodeInput) error {
|
||||
|
||||
switch pending.Type {
|
||||
case newagentmodel.PendingInteractionTypeAskUser:
|
||||
return handleInterruptAskUser(ctx, pending, conversationContext, emitter)
|
||||
return handleInterruptAskUser(ctx, runtimeState, input.PersistVisibleMessage, pending, conversationContext, emitter)
|
||||
case newagentmodel.PendingInteractionTypeConfirm:
|
||||
return handleInterruptConfirm(pending, emitter)
|
||||
default:
|
||||
@@ -70,6 +71,8 @@ func RunInterruptNode(ctx context.Context, input InterruptNodeInput) error {
|
||||
// 写入历史，然后结束。用户体验和正常对话一样 — 助手问了问题，停下来等回复。
|
||||
func handleInterruptAskUser(
|
||||
ctx context.Context,
|
||||
runtimeState *newagentmodel.AgentRuntimeState,
|
||||
persist newagentmodel.PersistVisibleMessageFunc,
|
||||
pending *newagentmodel.PendingInteraction,
|
||||
conversationContext *newagentmodel.ConversationContext,
|
||||
emitter *newagentstream.ChunkEmitter,
|
||||
@@ -89,7 +92,9 @@ func handleInterruptAskUser(
|
||||
}
|
||||
|
||||
// 写入对话历史，下一轮 resume 时 LLM 能看到这个上下文。
|
||||
conversationContext.AppendHistory(schema.AssistantMessage(text, nil))
|
||||
msg := schema.AssistantMessage(text, nil)
|
||||
conversationContext.AppendHistory(msg)
|
||||
persistVisibleAssistantMessage(ctx, persist, runtimeState.EnsureCommonState(), msg)
|
||||
|
||||
// 状态持久化已由 agent_nodes 层统一处理，此处不再需要自行存快照。
|
||||
|
||||
|
||||
121
backend/newAgent/node/llm_debug.go
Normal file
121
backend/newAgent/node/llm_debug.go
Normal file
@@ -0,0 +1,121 @@
|
||||
package newagentnode
|
||||
|
||||
import (
|
||||
"encoding/json"
|
||||
"fmt"
|
||||
"log"
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
)
|
||||
|
||||
// logNodeLLMContext 将某个节点即将送入 LLM 的完整消息上下文按统一格式打印到日志。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. 统一输出 stage / phase / chat / round，方便按一次请求内的多次 LLM 调用串联排查；
|
||||
// 2. 完整展开 messages，不做截断，保证问题复现时能直接对照 prompt 组装结果；
|
||||
// 3. 该函数只负责调试日志，不参与任何业务判断，也不修改上下文内容。
|
||||
func logNodeLLMContext(
|
||||
stage string,
|
||||
phase string,
|
||||
flowState *newagentmodel.CommonState,
|
||||
messages []*schema.Message,
|
||||
) {
|
||||
chatID := ""
|
||||
roundUsed := 0
|
||||
if flowState != nil {
|
||||
chatID = flowState.ConversationID
|
||||
roundUsed = flowState.RoundUsed
|
||||
}
|
||||
|
||||
log.Printf(
|
||||
"[DEBUG] %s LLM context begin phase=%s chat=%s round=%d message_count=%d\n%s\n[DEBUG] %s LLM context end phase=%s chat=%s round=%d",
|
||||
stage,
|
||||
strings.TrimSpace(phase),
|
||||
chatID,
|
||||
roundUsed,
|
||||
len(messages),
|
||||
formatLLMMessagesForDebug(messages),
|
||||
stage,
|
||||
strings.TrimSpace(phase),
|
||||
chatID,
|
||||
roundUsed,
|
||||
)
|
||||
}
|
||||
|
||||
// formatLLMMessagesForDebug 将本轮送入 LLM 的完整消息上下文展开成可读多行日志。
|
||||
//
|
||||
// 说明：
|
||||
// 1. 按消息索引逐条输出，便于和上游上下文构造步骤逐项对齐；
|
||||
// 2. 完整输出 content / reasoning_content / tool_calls / extra，不做截断；
|
||||
// 3. 仅用于调试打点，不参与业务决策。
|
||||
func formatLLMMessagesForDebug(messages []*schema.Message) string {
|
||||
if len(messages) == 0 {
|
||||
return "(empty messages)"
|
||||
}
|
||||
|
||||
var sb strings.Builder
|
||||
for i, msg := range messages {
|
||||
sb.WriteString(fmt.Sprintf("----- message[%d] -----\n", i))
|
||||
if msg == nil {
|
||||
sb.WriteString("role: <nil>\n\n")
|
||||
continue
|
||||
}
|
||||
|
||||
sb.WriteString(fmt.Sprintf("role: %s\n", msg.Role))
|
||||
|
||||
if strings.TrimSpace(msg.ToolCallID) != "" {
|
||||
sb.WriteString(fmt.Sprintf("tool_call_id: %s\n", msg.ToolCallID))
|
||||
}
|
||||
if strings.TrimSpace(msg.ToolName) != "" {
|
||||
sb.WriteString(fmt.Sprintf("tool_name: %s\n", msg.ToolName))
|
||||
}
|
||||
|
||||
if len(msg.ToolCalls) > 0 {
|
||||
sb.WriteString("tool_calls:\n")
|
||||
for j, call := range msg.ToolCalls {
|
||||
sb.WriteString(fmt.Sprintf(" - [%d] id=%s type=%s function=%s\n", j, call.ID, call.Type, call.Function.Name))
|
||||
sb.WriteString(" arguments:\n")
|
||||
sb.WriteString(indentMultilineForDebug(call.Function.Arguments, " "))
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
}
|
||||
|
||||
if strings.TrimSpace(msg.ReasoningContent) != "" {
|
||||
sb.WriteString("reasoning_content:\n")
|
||||
sb.WriteString(indentMultilineForDebug(msg.ReasoningContent, " "))
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
|
||||
sb.WriteString("content:\n")
|
||||
sb.WriteString(indentMultilineForDebug(msg.Content, " "))
|
||||
sb.WriteString("\n")
|
||||
|
||||
if len(msg.Extra) > 0 {
|
||||
sb.WriteString("extra:\n")
|
||||
raw, err := json.MarshalIndent(msg.Extra, "", " ")
|
||||
if err != nil {
|
||||
sb.WriteString(indentMultilineForDebug("<marshal_error>", " "))
|
||||
} else {
|
||||
sb.WriteString(indentMultilineForDebug(string(raw), " "))
|
||||
}
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
|
||||
sb.WriteString("\n")
|
||||
}
|
||||
return sb.String()
|
||||
}
|
||||
|
||||
// indentMultilineForDebug 为多行文本统一添加前缀缩进，避免日志折行后难以阅读。
|
||||
func indentMultilineForDebug(text, prefix string) string {
|
||||
if text == "" {
|
||||
return prefix + "<empty>"
|
||||
}
|
||||
lines := strings.Split(text, "\n")
|
||||
for i := range lines {
|
||||
lines[i] = prefix + lines[i]
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
@@ -28,14 +28,16 @@ const (
|
||||
|
||||
// PlanNodeInput 描述单轮规划节点执行所需的最小依赖。
|
||||
type PlanNodeInput struct {
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
UserInput string
|
||||
Client *infrallm.Client
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
ResumeNode string
|
||||
AlwaysExecute bool // true 时计划生成后自动确认，不进入 confirm 节点
|
||||
ThinkingEnabled bool // 是否开启 thinking，由 config.yaml 的 agent.thinking.plan 注入
|
||||
RuntimeState *newagentmodel.AgentRuntimeState
|
||||
ConversationContext *newagentmodel.ConversationContext
|
||||
UserInput string
|
||||
Client *infrallm.Client
|
||||
ChunkEmitter *newagentstream.ChunkEmitter
|
||||
ResumeNode string
|
||||
AlwaysExecute bool // true 时计划生成后自动确认，不进入 confirm 节点
|
||||
ThinkingEnabled bool // 是否开启 thinking，由 config.yaml 的 agent.thinking.plan 注入
|
||||
CompactionStore newagentmodel.CompactionStore // 上下文压缩持久化
|
||||
PersistVisibleMessage newagentmodel.PersistVisibleMessageFunc
|
||||
}
|
||||
|
||||
// RunPlanNode 执行一轮规划节点逻辑。
|
||||
@@ -68,6 +70,15 @@ func RunPlanNode(ctx context.Context, input PlanNodeInput) error {
|
||||
|
||||
// 2. 构造本轮规划输入。
|
||||
messages := newagentprompt.BuildPlanMessages(flowState, conversationContext, input.UserInput)
|
||||
messages = compactUnifiedMessagesIfNeeded(ctx, messages, UnifiedCompactInput{
|
||||
Client: input.Client,
|
||||
CompactionStore: input.CompactionStore,
|
||||
FlowState: flowState,
|
||||
Emitter: emitter,
|
||||
StageName: planStageName,
|
||||
StatusBlockID: planStatusBlockID,
|
||||
})
|
||||
logNodeLLMContext(planStageName, "planning", flowState, messages)
|
||||
|
||||
// 3. 单轮深度规划：由配置决定是否开启 thinking，不做 token 上限约束。
|
||||
decision, rawResult, err := infrallm.GenerateJSON[newagentmodel.PlanDecision](
|
||||
@@ -95,6 +106,7 @@ func RunPlanNode(ctx context.Context, input PlanNodeInput) error {
|
||||
|
||||
// 4. 若模型先对用户说了话，且不是 ask_user（ask_user 交给 interrupt 收口），则先以伪流式推送，再写回 history。
|
||||
if strings.TrimSpace(decision.Speak) != "" && decision.Action != newagentmodel.PlanActionAskUser {
|
||||
msg := schema.AssistantMessage(decision.Speak, nil)
|
||||
if err := emitter.EmitPseudoAssistantText(
|
||||
ctx,
|
||||
planSpeakBlockID,
|
||||
@@ -104,7 +116,8 @@ func RunPlanNode(ctx context.Context, input PlanNodeInput) error {
|
||||
); err != nil {
|
||||
return fmt.Errorf("规划文案推送失败: %w", err)
|
||||
}
|
||||
conversationContext.AppendHistory(schema.AssistantMessage(decision.Speak, nil))
|
||||
conversationContext.AppendHistory(msg)
|
||||
persistVisibleAssistantMessage(ctx, input.PersistVisibleMessage, flowState, msg)
|
||||
}
|
||||
|
||||
// 5. 按规划动作推进流程状态。
|
||||
@@ -139,6 +152,7 @@ func RunPlanNode(ctx context.Context, input PlanNodeInput) error {
|
||||
// 3. 推流后同步写入历史，确保后续 Execute 阶段的上下文也能看到这份计划。
|
||||
summary := strings.TrimSpace(buildPlanSummary(decision.PlanSteps))
|
||||
if summary != "" {
|
||||
msg := schema.AssistantMessage(summary, nil)
|
||||
if err := emitter.EmitPseudoAssistantText(
|
||||
ctx,
|
||||
planSummaryBlockID,
|
||||
@@ -148,7 +162,8 @@ func RunPlanNode(ctx context.Context, input PlanNodeInput) error {
|
||||
); err != nil {
|
||||
return fmt.Errorf("自动执行前计划摘要推送失败: %w", err)
|
||||
}
|
||||
conversationContext.AppendHistory(schema.AssistantMessage(summary, nil))
|
||||
conversationContext.AppendHistory(msg)
|
||||
persistVisibleAssistantMessage(ctx, input.PersistVisibleMessage, flowState, msg)
|
||||
}
|
||||
|
||||
flowState.ConfirmPlan()
|
||||
|
||||
301
backend/newAgent/node/unified_compact.go
Normal file
301
backend/newAgent/node/unified_compact.go
Normal file
@@ -0,0 +1,301 @@
|
||||
package newagentnode
|
||||
|
||||
import (
|
||||
"context"
|
||||
"encoding/json"
|
||||
"fmt"
|
||||
"log"
|
||||
|
||||
infrallm "github.com/LoveLosita/smartflow/backend/infra/llm"
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
newagentprompt "github.com/LoveLosita/smartflow/backend/newAgent/prompt"
|
||||
newagentstream "github.com/LoveLosita/smartflow/backend/newAgent/stream"
|
||||
"github.com/LoveLosita/smartflow/backend/pkg"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
)
|
||||
|
||||
// UnifiedCompactInput 是统一压缩入口的参数。
|
||||
//
|
||||
// 设计说明：
|
||||
// 1. 从 ExecuteNodeInput 中提取压缩所需的公共字段，消除对 Execute 的直接依赖；
|
||||
// 2. 各节点（Plan/Chat/Deliver）构造此参数时从自己的 NodeInput 中提取对应字段；
|
||||
// 3. StageName 和 StatusBlockID 用于区分日志来源和 SSE 状态推送。
|
||||
type UnifiedCompactInput struct {
|
||||
// Client 用于调用 LLM 压缩 msg1/msg2。
|
||||
Client *infrallm.Client
|
||||
// CompactionStore 用于持久化压缩摘要和 token 统计，为 nil 时跳过持久化。
|
||||
CompactionStore newagentmodel.CompactionStore
|
||||
// FlowState 提供 userID / chatID / roundUsed 等定位信息。
|
||||
FlowState *newagentmodel.CommonState
|
||||
// Emitter 用于推送压缩进度 SSE 事件。
|
||||
Emitter *newagentstream.ChunkEmitter
|
||||
// StageName 标识当前阶段（如 "execute"/"plan"/"chat"/"deliver"），用于日志和缓存 key。
|
||||
StageName string
|
||||
// StatusBlockID 是 SSE 状态推送的 block ID，各节点使用自己的 block ID。
|
||||
StatusBlockID string
|
||||
}
|
||||
|
||||
// compactUnifiedMessagesIfNeeded 检查统一消息结构的 token 预算，
|
||||
// 超限时对 msg1（历史对话）和 msg2（阶段工作区）执行 LLM 压缩。
|
||||
//
|
||||
// 消息布局约定（由 buildUnifiedStageMessages 返回）：
|
||||
//
|
||||
// [0] system — msg0: 系统规则 + 工具简表
|
||||
// [1] assistant — msg1: 历史对话上下文
|
||||
// [2] assistant — msg2: 阶段工作区（Execute=ReAct Loop，其余="暂无"）
|
||||
// [3] system — msg3: 阶段状态 + 记忆 + 指令
|
||||
//
|
||||
// 压缩策略：
|
||||
// 1. msg1 超过可用预算一半时触发 LLM 压缩（合并已有摘要 + 新内容）；
|
||||
// 2. msg1 压缩后仍超限，则对 msg2 也做 LLM 压缩；
|
||||
// 3. 压缩结果持久化到 CompactionStore，下一轮可复用摘要避免重复计算。
|
||||
func compactUnifiedMessagesIfNeeded(
|
||||
ctx context.Context,
|
||||
messages []*schema.Message,
|
||||
input UnifiedCompactInput,
|
||||
) []*schema.Message {
|
||||
if input.FlowState == nil {
|
||||
log.Printf("[COMPACT:%s] FlowState is nil, skip token stats refresh", input.StageName)
|
||||
return messages
|
||||
}
|
||||
|
||||
// 1. 非严格 4 段式时，退化成按角色汇总的统计，确保 context_token_stats 仍然刷新。
|
||||
if len(messages) != 4 {
|
||||
breakdown := estimateFallbackStageTokenBreakdown(messages)
|
||||
log.Printf(
|
||||
"[COMPACT:%s] fallback token stats refresh: total=%d budget=%d count=%d (msg0=%d msg1=%d msg2=%d msg3=%d)",
|
||||
input.StageName, breakdown.Total, breakdown.Budget, len(messages),
|
||||
breakdown.Msg0, breakdown.Msg1, breakdown.Msg2, breakdown.Msg3,
|
||||
)
|
||||
saveUnifiedTokenStats(ctx, input, breakdown)
|
||||
return messages
|
||||
}
|
||||
|
||||
// 2. 提取四条消息的文本内容。
|
||||
msg0 := messages[0].Content
|
||||
msg1 := messages[1].Content
|
||||
msg2 := messages[2].Content
|
||||
msg3 := messages[3].Content
|
||||
|
||||
// 3. Token 预算检查。
|
||||
breakdown, overBudget, needCompactMsg1, needCompactMsg2 := pkg.CheckStageTokenBudget(msg0, msg1, msg2, msg3)
|
||||
|
||||
log.Printf(
|
||||
"[COMPACT:%s] token budget check: total=%d budget=%d over=%v compactMsg1=%v compactMsg2=%v (msg0=%d msg1=%d msg2=%d msg3=%d)",
|
||||
input.StageName, breakdown.Total, breakdown.Budget, overBudget, needCompactMsg1, needCompactMsg2,
|
||||
breakdown.Msg0, breakdown.Msg1, breakdown.Msg2, breakdown.Msg3,
|
||||
)
|
||||
|
||||
if !overBudget {
|
||||
// 4. 未超限，记录 token 分布后直接返回。
|
||||
saveUnifiedTokenStats(ctx, input, breakdown)
|
||||
return messages
|
||||
}
|
||||
|
||||
// 5. msg1 压缩（历史对话 → LLM 摘要）。
|
||||
if needCompactMsg1 {
|
||||
msg1 = compactUnifiedMsg1(ctx, input, msg1)
|
||||
messages[1].Content = msg1
|
||||
// 压缩 msg1 后重算预算。
|
||||
breakdown = pkg.EstimateStageMessagesTokens(msg0, msg1, msg2, msg3)
|
||||
}
|
||||
|
||||
// 6. msg2 压缩（阶段工作区 → LLM 摘要）。
|
||||
if needCompactMsg2 || breakdown.Total > pkg.StageTokenBudget {
|
||||
msg2 = compactUnifiedMsg2(ctx, input, msg2)
|
||||
messages[2].Content = msg2
|
||||
breakdown = pkg.EstimateStageMessagesTokens(msg0, msg1, msg2, msg3)
|
||||
}
|
||||
|
||||
// 7. 记录最终 token 分布。
|
||||
saveUnifiedTokenStats(ctx, input, breakdown)
|
||||
|
||||
log.Printf(
|
||||
"[COMPACT:%s] after compaction: total=%d budget=%d (msg0=%d msg1=%d msg2=%d msg3=%d)",
|
||||
input.StageName, breakdown.Total, breakdown.Budget,
|
||||
breakdown.Msg0, breakdown.Msg1, breakdown.Msg2, breakdown.Msg3,
|
||||
)
|
||||
return messages
|
||||
}
|
||||
|
||||
// estimateFallbackStageTokenBreakdown 在非统一 4 段式场景下按消息角色做近似统计。
|
||||
//
|
||||
// 步骤说明：
|
||||
// 1. 先按消息类型汇总 token，保证总量准确；
|
||||
// 2. 再把最后一个 user 消息尽量视作 msg3，保留阶段指令语义；
|
||||
// 3. 其他历史内容归入 msg1 / msg2，确保上下文统计不会因为结构不标准而断更。
|
||||
func estimateFallbackStageTokenBreakdown(messages []*schema.Message) pkg.StageTokenBreakdown {
|
||||
breakdown := pkg.StageTokenBreakdown{Budget: pkg.StageTokenBudget}
|
||||
if len(messages) == 0 {
|
||||
return breakdown
|
||||
}
|
||||
|
||||
lastUserIndex := -1
|
||||
for i := len(messages) - 1; i >= 0; i-- {
|
||||
msg := messages[i]
|
||||
if msg == nil {
|
||||
continue
|
||||
}
|
||||
if msg.Role == schema.User {
|
||||
lastUserIndex = i
|
||||
break
|
||||
}
|
||||
}
|
||||
|
||||
for i, msg := range messages {
|
||||
if msg == nil {
|
||||
continue
|
||||
}
|
||||
tokens := pkg.EstimateMessageTokens(msg)
|
||||
breakdown.Total += tokens
|
||||
|
||||
switch msg.Role {
|
||||
case schema.System:
|
||||
breakdown.Msg0 += tokens
|
||||
case schema.User:
|
||||
if i == lastUserIndex {
|
||||
breakdown.Msg3 += tokens
|
||||
} else {
|
||||
breakdown.Msg1 += tokens
|
||||
}
|
||||
case schema.Tool:
|
||||
breakdown.Msg2 += tokens
|
||||
case schema.Assistant:
|
||||
if len(msg.ToolCalls) > 0 {
|
||||
breakdown.Msg2 += tokens
|
||||
} else {
|
||||
breakdown.Msg1 += tokens
|
||||
}
|
||||
default:
|
||||
breakdown.Msg1 += tokens
|
||||
}
|
||||
}
|
||||
|
||||
return breakdown
|
||||
}
|
||||
|
||||
// compactUnifiedMsg1 对 msg1（历史对话）执行 LLM 压缩。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. CompactionStore 为 nil 时跳过（测试环境 / 骨架期）；
|
||||
// 2. 先加载该阶段已有的压缩摘要，与当前 msg1 合并后调 LLM 压缩；
|
||||
// 3. 压缩失败时降级为原始文本，不中断主流程；
|
||||
// 4. 压缩成功后持久化新摘要，供下一轮复用。
|
||||
func compactUnifiedMsg1(
|
||||
ctx context.Context,
|
||||
input UnifiedCompactInput,
|
||||
msg1 string,
|
||||
) string {
|
||||
// 1. CompactionStore 为 nil 时无法加载/保存摘要，跳过压缩。
|
||||
if input.CompactionStore == nil {
|
||||
log.Printf("[COMPACT:%s] CompactionStore is nil, skip msg1 compaction", input.StageName)
|
||||
return msg1
|
||||
}
|
||||
|
||||
// 2. 加载该阶段已有的压缩摘要（可能为空）。
|
||||
existingSummary, _, err := input.CompactionStore.LoadStageCompaction(ctx, input.FlowState.UserID, input.FlowState.ConversationID, input.StageName)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT:%s] load existing compaction failed: %v, proceed without cache", input.StageName, err)
|
||||
}
|
||||
|
||||
// 3. SSE: 压缩开始。
|
||||
tokenBefore := pkg.EstimateTextTokens(msg1)
|
||||
_ = input.Emitter.EmitStatus(
|
||||
input.StatusBlockID, input.StageName, "context_compact_start",
|
||||
fmt.Sprintf("正在压缩对话历史（%d tokens）...", tokenBefore),
|
||||
false,
|
||||
)
|
||||
|
||||
// 4. 调用 LLM 压缩：将 msg1 全文 + 已有摘要合并为一份紧凑摘要。
|
||||
newSummary, err := newagentprompt.CompactMsg1(ctx, input.Client, msg1, existingSummary)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT:%s] compact msg1 failed: %v", input.StageName, err)
|
||||
_ = input.Emitter.EmitStatus(
|
||||
input.StatusBlockID, input.StageName, "context_compact_done",
|
||||
"对话历史压缩失败，使用原始文本",
|
||||
false,
|
||||
)
|
||||
return msg1
|
||||
}
|
||||
|
||||
// 5. SSE: 压缩完成。
|
||||
tokenAfter := pkg.EstimateTextTokens(newSummary)
|
||||
_ = input.Emitter.EmitStatus(
|
||||
input.StatusBlockID, input.StageName, "context_compact_done",
|
||||
fmt.Sprintf("对话历史已压缩：%d → %d tokens", tokenBefore, tokenAfter),
|
||||
false,
|
||||
)
|
||||
|
||||
// 6. 持久化压缩结果，下一轮可直接复用摘要。
|
||||
if err := input.CompactionStore.SaveStageCompaction(ctx, input.FlowState.UserID, input.FlowState.ConversationID, input.StageName, newSummary, input.FlowState.RoundUsed); err != nil {
|
||||
log.Printf("[COMPACT:%s] save compaction failed: %v", input.StageName, err)
|
||||
}
|
||||
|
||||
return newSummary
|
||||
}
|
||||
|
||||
// compactUnifiedMsg2 对 msg2（阶段工作区）执行 LLM 压缩。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. 非 Execute 阶段的 msg2 通常是"暂无"，压缩无意义但不会出错；
|
||||
// 2. Execute 阶段的 msg2 包含 ReAct loop 记录，压缩可显著节省 token；
|
||||
// 3. 压缩失败时降级为原始文本，不中断主流程。
|
||||
func compactUnifiedMsg2(
|
||||
ctx context.Context,
|
||||
input UnifiedCompactInput,
|
||||
msg2 string,
|
||||
) string {
|
||||
// 1. SSE: 压缩开始。
|
||||
tokenBefore := pkg.EstimateTextTokens(msg2)
|
||||
_ = input.Emitter.EmitStatus(
|
||||
input.StatusBlockID, input.StageName, "context_compact_start",
|
||||
fmt.Sprintf("正在压缩执行记录（%d tokens）...", tokenBefore),
|
||||
false,
|
||||
)
|
||||
|
||||
// 2. 调用 LLM 压缩。
|
||||
compressed, err := newagentprompt.CompactMsg2(ctx, input.Client, msg2)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT:%s] compact msg2 failed: %v", input.StageName, err)
|
||||
_ = input.Emitter.EmitStatus(
|
||||
input.StatusBlockID, input.StageName, "context_compact_done",
|
||||
"执行记录压缩失败，使用原始文本",
|
||||
false,
|
||||
)
|
||||
return msg2
|
||||
}
|
||||
|
||||
// 3. SSE: 压缩完成。
|
||||
tokenAfter := pkg.EstimateTextTokens(compressed)
|
||||
_ = input.Emitter.EmitStatus(
|
||||
input.StatusBlockID, input.StageName, "context_compact_done",
|
||||
fmt.Sprintf("执行记录已压缩：%d → %d tokens", tokenBefore, tokenAfter),
|
||||
false,
|
||||
)
|
||||
|
||||
return compressed
|
||||
}
|
||||
|
||||
// saveUnifiedTokenStats 持久化当前 token 分布到 DB。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. CompactionStore 为 nil 时跳过（测试环境 / 骨架期）；
|
||||
// 2. 序列化失败只记日志，不中断主流程；
|
||||
// 3. 写入失败只记日志，不中断主流程。
|
||||
func saveUnifiedTokenStats(
|
||||
ctx context.Context,
|
||||
input UnifiedCompactInput,
|
||||
breakdown pkg.StageTokenBreakdown,
|
||||
) {
|
||||
if input.CompactionStore == nil || input.FlowState == nil {
|
||||
return
|
||||
}
|
||||
statsJSON, err := json.Marshal(breakdown)
|
||||
if err != nil {
|
||||
log.Printf("[COMPACT:%s] marshal token stats failed: %v", input.StageName, err)
|
||||
return
|
||||
}
|
||||
if err := input.CompactionStore.SaveContextTokenStats(ctx, input.FlowState.UserID, input.FlowState.ConversationID, string(statsJSON)); err != nil {
|
||||
log.Printf("[COMPACT:%s] save token stats failed: %v", input.StageName, err)
|
||||
}
|
||||
}
|
||||
37
backend/newAgent/node/visible_message.go
Normal file
37
backend/newAgent/node/visible_message.go
Normal file
@@ -0,0 +1,37 @@
|
||||
package newagentnode
|
||||
|
||||
import (
|
||||
"context"
|
||||
"log"
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
)
|
||||
|
||||
// persistVisibleAssistantMessage 负责把“真正要展示给用户”的 assistant 文本交给 service 层持久化。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 只处理可见的 assistant 消息，不处理内部纠错提示、工具调用结果和纯状态文案；
|
||||
// 2. 持久化失败只记日志，不反向中断节点主流程，避免“已经对外输出但后端补写失败”时把用户请求打断；
|
||||
// 3. 具体的 Redis / MySQL / 乐观缓存写入由 service 回调统一完成。
|
||||
func persistVisibleAssistantMessage(
|
||||
ctx context.Context,
|
||||
persist newagentmodel.PersistVisibleMessageFunc,
|
||||
state *newagentmodel.CommonState,
|
||||
msg *schema.Message,
|
||||
) {
|
||||
if persist == nil || state == nil || msg == nil {
|
||||
return
|
||||
}
|
||||
|
||||
role := strings.TrimSpace(string(msg.Role))
|
||||
content := strings.TrimSpace(msg.Content)
|
||||
if role != string(schema.Assistant) || content == "" {
|
||||
return
|
||||
}
|
||||
|
||||
if err := persist(ctx, state, msg); err != nil {
|
||||
log.Printf("[WARN] persist visible assistant message failed chat=%s phase=%s err=%v", state.ConversationID, state.Phase, err)
|
||||
}
|
||||
}
|
||||
@@ -25,7 +25,7 @@ const chatRoutingSystemPrompt = `
|
||||
- route=direct_reply 时，控制码后的可见内容应直接回应用户问题，而不是先讲能力边界。
|
||||
- route=deep_answer 时，只输出控制码即可，不要补“让我想想”“这是个好问题”之类的占位话术。
|
||||
|
||||
粗排判断：当用户意图包含"批量安排/排课/把任务类排进日程"，且上下文中有任务类 ID 时，设置 rough_build=true。
|
||||
粗排判断：当用户意图包含"批量安排/排课/把任务类排进日程"等批量调度需求时，可设置 rough_build=true；后端会结合真实请求范围决定是否真正进入粗排。
|
||||
二次粗排约束（强约束）：
|
||||
- 若上下文已出现 rough_build_done，且用户未明确要求"重新粗排/从头重排"，必须设置 rough_build=false。
|
||||
- "移动/微调/优化/均匀化/调顺序"等请求默认视为 refine，不得再次触发 rough build。
|
||||
@@ -83,40 +83,25 @@ func BuildChatRoutingSystemPrompt() string {
|
||||
|
||||
// BuildChatRoutingMessages 组装路由阶段的 messages。
|
||||
func BuildChatRoutingMessages(ctx *newagentmodel.ConversationContext, userInput string, state *newagentmodel.CommonState, nonce string) []*schema.Message {
|
||||
return buildStageMessages(
|
||||
BuildChatRoutingSystemPrompt(),
|
||||
return buildUnifiedStageMessages(
|
||||
ctx,
|
||||
BuildChatRoutingUserPrompt(ctx, userInput, state, nonce),
|
||||
StageMessagesConfig{
|
||||
SystemPrompt: BuildChatRoutingSystemPrompt(),
|
||||
Msg1Content: buildChatConversationMessage(ctx),
|
||||
Msg2Content: buildChatRoutingWorkspace(ctx),
|
||||
Msg3Suffix: BuildChatRoutingUserPrompt(userInput, nonce),
|
||||
Msg3Role: schema.User,
|
||||
},
|
||||
)
|
||||
}
|
||||
|
||||
// BuildChatRoutingUserPrompt 构造路由阶段的用户提示词。
|
||||
func BuildChatRoutingUserPrompt(ctx *newagentmodel.ConversationContext, userInput string, state *newagentmodel.CommonState, nonce string) string {
|
||||
func BuildChatRoutingUserPrompt(userInput string, nonce string) string {
|
||||
var sb strings.Builder
|
||||
|
||||
sb.WriteString(fmt.Sprintf("nonce=%s\n", nonce))
|
||||
sb.WriteString(fmt.Sprintf("当前时间=%s\n", time.Now().In(time.Local).Format("2006-01-02 15:04")))
|
||||
sb.WriteString("\n请判断用户本轮意图的复杂度，选择最合适的路由，并输出控制码和对应内容。\n")
|
||||
|
||||
// 注入任务类上下文（供粗排判断参考）。
|
||||
if state != nil && len(state.TaskClassIDs) > 0 {
|
||||
parts := make([]string, len(state.TaskClassIDs))
|
||||
for i, id := range state.TaskClassIDs {
|
||||
parts[i] = fmt.Sprintf("%d", id)
|
||||
}
|
||||
sb.WriteString(fmt.Sprintf("\n本次请求涉及的任务类 ID：[%s]\n", strings.Join(parts, ", ")))
|
||||
}
|
||||
|
||||
if state != nil && len(state.TaskClasses) > 0 {
|
||||
sb.WriteString("任务类约束：\n")
|
||||
for _, tc := range state.TaskClasses {
|
||||
line := fmt.Sprintf("- [ID=%d] %s：策略=%s，总时段预算=%d", tc.ID, tc.Name, tc.Strategy, tc.TotalSlots)
|
||||
if tc.StartDate != "" || tc.EndDate != "" {
|
||||
line += fmt.Sprintf("，日期范围=%s ~ %s", tc.StartDate, tc.EndDate)
|
||||
}
|
||||
sb.WriteString(line + "\n")
|
||||
}
|
||||
}
|
||||
sb.WriteString("\n请基于最近真实对话和本轮输入选择最合适的路由，并严格按系统约定输出控制码。\n")
|
||||
|
||||
trimmedInput := strings.TrimSpace(userInput)
|
||||
if trimmedInput != "" {
|
||||
@@ -146,10 +131,23 @@ func BuildDeepAnswerSystemPrompt() string {
|
||||
}
|
||||
|
||||
// BuildDeepAnswerMessages 组装深度回答阶段的 messages。
|
||||
func BuildDeepAnswerMessages(ctx *newagentmodel.ConversationContext, userInput string) []*schema.Message {
|
||||
return buildStageMessages(
|
||||
BuildDeepAnswerSystemPrompt(),
|
||||
func BuildDeepAnswerMessages(state *newagentmodel.CommonState, ctx *newagentmodel.ConversationContext, userInput string) []*schema.Message {
|
||||
return buildUnifiedStageMessages(
|
||||
ctx,
|
||||
userInput,
|
||||
StageMessagesConfig{
|
||||
SystemPrompt: BuildDeepAnswerSystemPrompt(),
|
||||
Msg1Content: buildChatConversationMessage(ctx),
|
||||
Msg2Content: buildDeepAnswerWorkspace(),
|
||||
Msg3Suffix: buildDeepAnswerUserPrompt(userInput),
|
||||
Msg3Role: schema.User,
|
||||
},
|
||||
)
|
||||
}
|
||||
|
||||
func buildDeepAnswerUserPrompt(userInput string) string {
|
||||
trimmedInput := strings.TrimSpace(userInput)
|
||||
if trimmedInput != "" {
|
||||
return trimmedInput
|
||||
}
|
||||
return "请直接回答用户刚才的问题。"
|
||||
}
|
||||
|
||||
33
backend/newAgent/prompt/chat_context.go
Normal file
33
backend/newAgent/prompt/chat_context.go
Normal file
@@ -0,0 +1,33 @@
|
||||
package newagentprompt
|
||||
|
||||
import (
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
)
|
||||
|
||||
// buildChatConversationMessage 生成 chat / deep_answer 共用的真实对话视图。
|
||||
func buildChatConversationMessage(ctx *newagentmodel.ConversationContext) string {
|
||||
return buildConversationHistoryMessage(ctx, "真实对话记录")
|
||||
}
|
||||
|
||||
// buildChatRoutingWorkspace 渲染 chat 路由节点的轻量补充区。
|
||||
//
|
||||
// 设计说明：
|
||||
// 1. chat 只保留与路由判断直接相关的最小流程标记；
|
||||
// 2. rough_build_done 仍需显式暴露，否则路由层会丢掉“不要重复粗排”的关键信号；
|
||||
// 3. 不再展示轮次、阶段锚点、ReAct 摘要等 execute 专属信息。
|
||||
func buildChatRoutingWorkspace(ctx *newagentmodel.ConversationContext) string {
|
||||
lines := []string{"路由补充："}
|
||||
if hasExecuteRoughBuildDone(ctx) {
|
||||
lines = append(lines, "- 已存在 rough_build_done；除非用户明确要求重新粗排，否则不要再次触发 rough_build。")
|
||||
} else {
|
||||
lines = append(lines, "- 暂无额外流程标记。")
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// buildDeepAnswerWorkspace 渲染 deep_answer 节点的轻量工作区。
|
||||
func buildDeepAnswerWorkspace() string {
|
||||
return "回答补充：请直接延续最近对话，聚焦回答用户本轮问题。"
|
||||
}
|
||||
37
backend/newAgent/prompt/conversation_view.go
Normal file
37
backend/newAgent/prompt/conversation_view.go
Normal file
@@ -0,0 +1,37 @@
|
||||
package newagentprompt
|
||||
|
||||
import (
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
)
|
||||
|
||||
// buildConversationHistoryMessage 将“真实对话流”渲染成节点可直接复用的 msg1。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 只负责把 user + assistant speak 组织成稳定文本；
|
||||
// 2. 不拼接 tool_call / tool observation，这些不属于“真实对话”；
|
||||
// 3. 不做长度裁剪，长度预算交给统一压缩层处理。
|
||||
func buildConversationHistoryMessage(ctx *newagentmodel.ConversationContext, title string) string {
|
||||
title = strings.TrimSpace(title)
|
||||
if title == "" {
|
||||
title = "真实对话记录"
|
||||
}
|
||||
|
||||
lines := []string{title + "："}
|
||||
if ctx == nil {
|
||||
lines = append(lines, "暂无。")
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
turns := CollectConversationTurns(ctx.HistorySnapshot())
|
||||
if len(turns) == 0 {
|
||||
lines = append(lines, "暂无。")
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
for _, turn := range turns {
|
||||
lines = append(lines, turn.Role+": \""+turn.Content+"\"")
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
@@ -14,16 +14,16 @@ const deliverSystemPrompt = `
|
||||
|
||||
请遵守以下规则：
|
||||
1. 只基于已有历史和计划状态生成总结，不要编造未执行的操作。
|
||||
2. 如果所有步骤都已完成，简要总结每一步的成果。
|
||||
3. 如果是因轮次耗尽提前结束，如实告知用户当前进度和未完成的部分。
|
||||
4. 使用自然、友好的语气，不要机械地罗列步骤。
|
||||
5. 如果用户后续可能需要继续操作，给出简短的建议。
|
||||
6. 只输出总结文本，不要输出 JSON，不要输出 markdown 标题。
|
||||
2. 如果所有步骤都已完成，请自然概括每一步的主要成果。
|
||||
3. 如果流程因轮次耗尽或主动终止而提前结束，请如实说明当前进度与未完成部分。
|
||||
4. 使用自然、友好的语气，不要机械罗列工具过程。
|
||||
5. 如果用户后续还需要继续操作，可以给出一句简短建议。
|
||||
6. 只输出总结文本，不要输出 JSON，也不要输出 markdown 标题。
|
||||
|
||||
你会看到：
|
||||
- 原始计划步骤及完成判定
|
||||
- 当前执行进度
|
||||
- 执行阶段的对话历史
|
||||
- 原始计划步骤及完成进度
|
||||
- 最近真实对话
|
||||
- 当前流程的收口状态
|
||||
`
|
||||
|
||||
// BuildDeliverSystemPrompt 返回交付阶段系统提示词。
|
||||
@@ -31,37 +31,52 @@ func BuildDeliverSystemPrompt() string {
|
||||
return strings.TrimSpace(deliverSystemPrompt)
|
||||
}
|
||||
|
||||
// BuildDeliverMessages 组装交付阶段的 messages。
|
||||
// BuildDeliverMessages 组装交付阶段 messages。
|
||||
func BuildDeliverMessages(state *newagentmodel.CommonState, ctx *newagentmodel.ConversationContext) []*schema.Message {
|
||||
return buildStageMessages(
|
||||
BuildDeliverSystemPrompt(),
|
||||
roughBuildPrefix := buildDeliverRoughBuildPrefix(ctx, state)
|
||||
return buildUnifiedStageMessages(
|
||||
ctx,
|
||||
BuildDeliverUserPrompt(state),
|
||||
StageMessagesConfig{
|
||||
SystemPrompt: BuildDeliverSystemPrompt(),
|
||||
Msg1Content: buildDeliverConversationMessage(ctx),
|
||||
Msg2Content: buildDeliverWorkspace(state),
|
||||
Msg3Prefix: roughBuildPrefix,
|
||||
Msg3Suffix: BuildDeliverUserPrompt(state, ctx),
|
||||
Msg3Role: schema.User,
|
||||
},
|
||||
)
|
||||
}
|
||||
|
||||
// BuildDeliverUserPrompt 构造交付阶段的用户提示词。
|
||||
func BuildDeliverUserPrompt(state *newagentmodel.CommonState) string {
|
||||
func BuildDeliverUserPrompt(state *newagentmodel.CommonState, ctx *newagentmodel.ConversationContext) string {
|
||||
var sb strings.Builder
|
||||
|
||||
sb.WriteString("请为当前任务生成完成总结。\n")
|
||||
sb.WriteString(renderStateSummary(state))
|
||||
sb.WriteString("\n")
|
||||
sb.WriteString("请基于最近对话和交付工作区，生成一段自然、诚实的完成总结。\n")
|
||||
|
||||
if state == nil || !state.HasPlan() {
|
||||
sb.WriteString("当前没有正式计划，请基于对话历史简要总结本次交互。\n")
|
||||
if hasExecuteRoughBuildDone(ctx) {
|
||||
sb.WriteString("当前没有正式计划，但本轮已经完成粗排，请结合粗排补充和任务类详情总结粗排结果，不要把它说成正式完结。\n")
|
||||
} else {
|
||||
sb.WriteString("当前没有正式计划，请只概括本次互动，不要编造成果。\n")
|
||||
}
|
||||
return strings.TrimSpace(sb.String())
|
||||
}
|
||||
|
||||
current, total := state.PlanProgress()
|
||||
exhausted := state.Exhausted()
|
||||
completed := countCompletedPlanSteps(state)
|
||||
total := len(state.PlanSteps)
|
||||
|
||||
if exhausted {
|
||||
sb.WriteString(fmt.Sprintf("注意：任务因轮次耗尽提前结束，当前进度 %d/%d。\n", current, total))
|
||||
sb.WriteString("请如实说明已完成和未完成的部分，并建议用户如何继续。\n")
|
||||
} else {
|
||||
sb.WriteString("所有计划步骤已执行完毕，请总结整体成果。\n")
|
||||
if state.IsExhaustedTerminal() {
|
||||
sb.WriteString(fmt.Sprintf("注意：任务因轮次耗尽提前结束，当前已完成 %d/%d 步。\n", completed, total))
|
||||
sb.WriteString("请如实说明已完成与未完成的部分，并给出一句继续建议。\n")
|
||||
return strings.TrimSpace(sb.String())
|
||||
}
|
||||
|
||||
if state.IsAborted() {
|
||||
sb.WriteString(fmt.Sprintf("注意：流程已被主动终止，当前已完成 %d/%d 步。\n", completed, total))
|
||||
sb.WriteString("请如实说明停在何处，以及用户若想继续应如何衔接。\n")
|
||||
return strings.TrimSpace(sb.String())
|
||||
}
|
||||
|
||||
sb.WriteString("若计划已正常完成，请概括整体成果；若仍有未完成步骤，也必须如实说明。\n")
|
||||
return strings.TrimSpace(sb.String())
|
||||
}
|
||||
|
||||
137
backend/newAgent/prompt/deliver_context.go
Normal file
137
backend/newAgent/prompt/deliver_context.go
Normal file
@@ -0,0 +1,137 @@
|
||||
package newagentprompt
|
||||
|
||||
import (
|
||||
"fmt"
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
)
|
||||
|
||||
// buildDeliverConversationMessage 生成 deliver 节点看到的真实对话视图。
|
||||
func buildDeliverConversationMessage(ctx *newagentmodel.ConversationContext) string {
|
||||
return buildConversationHistoryMessage(ctx, "执行对话记录")
|
||||
}
|
||||
|
||||
// buildDeliverRoughBuildPrefix 构造 deliver 在“粗排已完成”场景下的专属前缀。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只负责把粗排相关的任务类信息补进 msg3 前缀，不改写交付总结本身；
|
||||
// 2. 只有在上下文里明确存在 rough_build_done 时才注入，避免普通交付场景被额外信息污染；
|
||||
// 3. 这段前缀用于补齐第一次粗排没有正式计划时的任务类详情，优先让 deliver 看到 task_class_ids 和任务类约束。
|
||||
func buildDeliverRoughBuildPrefix(ctx *newagentmodel.ConversationContext, state *newagentmodel.CommonState) string {
|
||||
if !hasExecuteRoughBuildDone(ctx) {
|
||||
return ""
|
||||
}
|
||||
|
||||
lines := []string{
|
||||
"粗排补充信息：",
|
||||
"- 本轮已经完成粗排，相关任务类已进入 suggested/existing，不要把它们说成正式计划。",
|
||||
}
|
||||
|
||||
if taskClassIDs := renderPlanTaskClassIDs(state); taskClassIDs != "" {
|
||||
lines = append(lines, "- "+taskClassIDs)
|
||||
}
|
||||
if taskClassMeta := renderPlanTaskClassMeta(state); taskClassMeta != "" {
|
||||
lines = append(lines, "任务类详情：")
|
||||
lines = append(lines, taskClassMeta)
|
||||
}
|
||||
|
||||
if state == nil || !state.HasPlan() {
|
||||
lines = append(lines, "- 当前没有正式计划，请把这批任务类的粗排结果作为总结重点。")
|
||||
}
|
||||
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// buildDeliverWorkspace 渲染 deliver 节点自己的结果视图。
|
||||
//
|
||||
// 设计说明：
|
||||
// 1. deliver 只需要结果态信息：计划简表、完成进度、收口状态；
|
||||
// 2. 不再注入工具目录、任务类约束、ReAct 摘要等过程噪声；
|
||||
// 3. 没有正式计划时，明确退回“只基于对话做总结”。
|
||||
func buildDeliverWorkspace(state *newagentmodel.CommonState) string {
|
||||
lines := []string{"交付工作区："}
|
||||
if state == nil {
|
||||
lines = append(lines, "- 当前缺少流程状态，请仅基于最近对话做诚实总结。")
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
lines = append(lines, renderDeliverTerminalSummary(state))
|
||||
if !state.HasPlan() {
|
||||
lines = append(lines, "- 当前没有正式计划，请只概括本次互动。")
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
total := len(state.PlanSteps)
|
||||
completed := countCompletedPlanSteps(state)
|
||||
lines = append(lines, fmt.Sprintf("- 计划进度：已完成 %d/%d 步。", completed, total))
|
||||
lines = append(lines, "计划步骤：")
|
||||
lines = append(lines, renderDeliverStepOutline(state, completed))
|
||||
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// renderDeliverTerminalSummary 返回 deliver 节点需要知道的收口状态。
|
||||
func renderDeliverTerminalSummary(state *newagentmodel.CommonState) string {
|
||||
if state == nil || !state.HasTerminalOutcome() || state.TerminalOutcome == nil {
|
||||
return "- 当前没有正式终止结果，请按最近对话和计划进度自然总结。"
|
||||
}
|
||||
|
||||
outcome := state.TerminalOutcome
|
||||
line := fmt.Sprintf("- 收口状态：%s", outcome.Status)
|
||||
if stage := strings.TrimSpace(outcome.Stage); stage != "" {
|
||||
line += fmt.Sprintf("；阶段：%s", stage)
|
||||
}
|
||||
if msg := strings.TrimSpace(outcome.UserMessage); msg != "" {
|
||||
line += fmt.Sprintf("；用户提示：%s", msg)
|
||||
}
|
||||
return line
|
||||
}
|
||||
|
||||
// renderDeliverStepOutline 生成 deliver 节点使用的步骤简表。
|
||||
func renderDeliverStepOutline(state *newagentmodel.CommonState, completed int) string {
|
||||
if state == nil || len(state.PlanSteps) == 0 {
|
||||
return "- 暂无。"
|
||||
}
|
||||
|
||||
lines := make([]string, 0, len(state.PlanSteps))
|
||||
for i, step := range state.PlanSteps {
|
||||
status := "未完成"
|
||||
if i < completed {
|
||||
status = "已完成"
|
||||
}
|
||||
|
||||
content := strings.TrimSpace(step.Content)
|
||||
if content == "" {
|
||||
content = "（步骤正文为空）"
|
||||
}
|
||||
line := fmt.Sprintf("%d. [%s] %s", i+1, status, content)
|
||||
if doneWhen := strings.TrimSpace(step.DoneWhen); doneWhen != "" {
|
||||
line += fmt.Sprintf(" | 完成判定：%s", doneWhen)
|
||||
}
|
||||
lines = append(lines, line)
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// countCompletedPlanSteps 统计当前已经完成的计划步骤数。
|
||||
func countCompletedPlanSteps(state *newagentmodel.CommonState) int {
|
||||
if state == nil {
|
||||
return 0
|
||||
}
|
||||
|
||||
total := len(state.PlanSteps)
|
||||
if total == 0 {
|
||||
return 0
|
||||
}
|
||||
if state.CurrentStep <= 0 {
|
||||
if state.IsCompleted() {
|
||||
return total
|
||||
}
|
||||
return 0
|
||||
}
|
||||
if state.CurrentStep >= total {
|
||||
return total
|
||||
}
|
||||
return state.CurrentStep
|
||||
}
|
||||
@@ -12,20 +12,11 @@ import (
|
||||
)
|
||||
|
||||
const (
|
||||
// executeHistoryKindKey 用于在 history 中打运行态标记，供 prompt 分层识别。
|
||||
// 说明：loop_closed / step_advanced 等边界标记仍由节点层写入，但 prompt 层已不再消费它们——
|
||||
// 因为 msg1/msg2 已经按"真实对话流 + 当前活跃 ReAct 记录"重构，不再做 msg2→msg1 的归档搬运。
|
||||
executeHistoryKindKey = "newagent_history_kind"
|
||||
executeHistoryKindCorrectionUser = "llm_correction_prompt"
|
||||
executeHistoryKindLoopClosed = "execute_loop_closed"
|
||||
executeHistoryKindStepAdvanced = "execute_step_advanced"
|
||||
|
||||
// executeLoopWindowLimit 控制当轮 ReAct Loop 窗口最多保留多少条记录。
|
||||
executeLoopWindowLimit = 8
|
||||
|
||||
// executeTrimmedObservationText 是重复工具压缩后的 observation 占位文案。
|
||||
executeTrimmedObservationText = "当前工具调用结果已经被使用过，当前无需使用，为节省上下文空间，已折叠"
|
||||
|
||||
// executeConversationTurnLimit 控制 msg1 注入的最大对话轮数（user + assistant speak）。
|
||||
// 超出时保留最近的条目，早期部分由 ReAct 摘要兜底。
|
||||
executeConversationTurnLimit = 30
|
||||
)
|
||||
|
||||
type executeToolSchemaDoc struct {
|
||||
@@ -40,8 +31,6 @@ type executeLoopRecord struct {
|
||||
Observation string
|
||||
}
|
||||
|
||||
const executeMessage1MaxRunes = 1400
|
||||
|
||||
// buildExecuteStageMessages 组装 execute 阶段 4 条消息骨架。
|
||||
//
|
||||
// 消息结构（固定）：
|
||||
@@ -82,87 +71,24 @@ func buildExecuteMessage0(stageSystemPrompt string, ctx *newagentmodel.Conversat
|
||||
return base + "\n\n" + toolCatalog
|
||||
}
|
||||
|
||||
// splitExecuteLoopRecordsByBoundary 按已收口标记拆分归档/活跃 ReAct 记录。
|
||||
//
|
||||
// 规则：
|
||||
// 1. 标记之前的记录归档到 msg1；
|
||||
// 2. 标记之后的记录作为活跃 loop 进入 msg2；
|
||||
// 3. 若没有标记，则全部视为活跃记录（兼容旧会话快照）。
|
||||
func splitExecuteLoopRecordsByBoundary(history []*schema.Message) (archived []executeLoopRecord, active []executeLoopRecord) {
|
||||
if len(history) == 0 {
|
||||
return nil, nil
|
||||
}
|
||||
|
||||
boundary := findLatestExecuteBoundaryMarker(history)
|
||||
if boundary < 0 {
|
||||
return nil, collectExecuteLoopRecords(history)
|
||||
}
|
||||
|
||||
if boundary > 0 {
|
||||
archived = collectExecuteLoopRecords(history[:boundary])
|
||||
}
|
||||
if boundary+1 < len(history) {
|
||||
active = collectExecuteLoopRecords(history[boundary+1:])
|
||||
}
|
||||
return archived, active
|
||||
}
|
||||
|
||||
func findLatestExecuteBoundaryMarker(history []*schema.Message) int {
|
||||
for i := len(history) - 1; i >= 0; i-- {
|
||||
msg := history[i]
|
||||
if msg == nil || msg.Extra == nil {
|
||||
continue
|
||||
}
|
||||
kind, ok := msg.Extra[executeHistoryKindKey].(string)
|
||||
if !ok {
|
||||
continue
|
||||
}
|
||||
switch strings.TrimSpace(kind) {
|
||||
case executeHistoryKindLoopClosed, executeHistoryKindStepAdvanced:
|
||||
return i
|
||||
}
|
||||
}
|
||||
return -1
|
||||
}
|
||||
|
||||
func trimExecuteMessage1ByBudget(content string) string {
|
||||
content = strings.TrimSpace(content)
|
||||
if content == "" {
|
||||
return ""
|
||||
}
|
||||
runes := []rune(content)
|
||||
if len(runes) <= executeMessage1MaxRunes {
|
||||
return content
|
||||
}
|
||||
if executeMessage1MaxRunes <= 3 {
|
||||
return string(runes[:executeMessage1MaxRunes])
|
||||
}
|
||||
return string(runes[:executeMessage1MaxRunes-3]) + "..."
|
||||
}
|
||||
|
||||
// buildExecuteMessage1V3 负责把真实对话流 + 上一轮 loop 归档并入 msg1，并统一做长度裁剪。
|
||||
// buildExecuteMessage1V3 只渲染"真实对话流 + 阶段锚点"。
|
||||
//
|
||||
// 改造说明：
|
||||
// 1. msg1 从人工提炼的摘要变为真实对话流，只注入 user + assistant speak；
|
||||
// 2. tool_call / observation 不在 msg1 中重复（已由 msg2 承载）；
|
||||
// 3. 超出 executeConversationTurnLimit 的早期对话不注入，由 ReAct 摘要兜底。
|
||||
// 1. msg1 只保留 user + assistant speak 组成的真实对话历史，全量注入；
|
||||
// 2. tool_call / observation 一律由 msg2 承载，这里不再重复；
|
||||
// 3. 不再从历史中"归档"上一轮 ReAct 结果到 msg1——归档搬运逻辑已随 splitExecuteLoopRecordsByBoundary 一并移除；
|
||||
// 4. token 预算由统一压缩层兜底，prompt 层不做提前裁剪。
|
||||
func buildExecuteMessage1V3(ctx *newagentmodel.ConversationContext) string {
|
||||
lines := []string{"历史上下文："}
|
||||
if ctx == nil {
|
||||
lines = append(lines,
|
||||
"- 对话历史：暂无。",
|
||||
"- 阶段锚点：按当前工具事实推进执行。",
|
||||
"- 历史归档 ReAct 摘要：暂无。",
|
||||
"- 历史归档 ReAct 窗口：暂无。",
|
||||
"- 当前循环早期摘要：暂无。",
|
||||
)
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
history := ctx.HistorySnapshot()
|
||||
|
||||
// 注入真实对话流（user + assistant speak），全量放入，不再限制轮数和单条长度。
|
||||
turns := collectExecuteConversationTurns(history)
|
||||
turns := collectExecuteConversationTurns(ctx.HistorySnapshot())
|
||||
if len(turns) == 0 {
|
||||
lines = append(lines, "- 对话历史：暂无。")
|
||||
} else {
|
||||
@@ -180,16 +106,15 @@ func buildExecuteMessage1V3(ctx *newagentmodel.ConversationContext) string {
|
||||
lines = append(lines, "- 阶段锚点：按当前工具事实推进，不做无依据操作。")
|
||||
}
|
||||
|
||||
archivedLoops, activeLoops := splitExecuteLoopRecordsByBoundary(history)
|
||||
lines = append(lines, "- 历史归档 ReAct 摘要："+buildEarlyExecuteReactSummary(archivedLoops, executeLoopWindowLimit))
|
||||
lines = append(lines, renderArchivedExecuteLoopWindowForMessage1V3(archivedLoops))
|
||||
lines = append(lines, "- 当前循环早期摘要："+buildEarlyExecuteReactSummary(activeLoops, executeLoopWindowLimit))
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// buildExecuteMessage2V3 承载当前活跃 loop 的全部记录。
|
||||
// 若是新一轮刚开始（活跃 loop 为空），明确返回已清空状态。
|
||||
// 不再限制窗口大小，token 预算由 execute 层统一管理。
|
||||
// buildExecuteMessage2V3 承载当前会话中全部 ReAct Loop 记录。
|
||||
//
|
||||
// 改造说明：
|
||||
// 1. 不再按 execute_loop_closed / execute_step_advanced 边界切分"归档/活跃"两段；
|
||||
// 2. 直接从 history 提取全部 assistant tool_call + 对应 observation 作为当前 Loop 视图；
|
||||
// 3. 新一轮刚开始（尚未产生 tool_call）时返回明确占位，方便模型识别"干净起点"。
|
||||
func buildExecuteMessage2V3(ctx *newagentmodel.ConversationContext) string {
|
||||
lines := []string{"当轮 ReAct Loop 记录："}
|
||||
if ctx == nil {
|
||||
@@ -197,31 +122,13 @@ func buildExecuteMessage2V3(ctx *newagentmodel.ConversationContext) string {
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
_, activeLoops := splitExecuteLoopRecordsByBoundary(ctx.HistorySnapshot())
|
||||
if len(activeLoops) == 0 {
|
||||
loops := collectExecuteLoopRecords(ctx.HistorySnapshot())
|
||||
if len(loops) == 0 {
|
||||
lines = append(lines, "- 已清空（新一轮 loop 准备中）。")
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// 全量放入，不再限制窗口大小
|
||||
for i, loop := range activeLoops {
|
||||
lines = append(lines, fmt.Sprintf("%d) thought/reason：%s", i+1, loop.Thought))
|
||||
lines = append(lines, fmt.Sprintf(" tool_call：%s", renderExecuteToolCallText(loop.ToolName, loop.ToolArgs)))
|
||||
lines = append(lines, fmt.Sprintf(" observation：%s", loop.Observation))
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
func renderArchivedExecuteLoopWindowForMessage1V3(records []executeLoopRecord) string {
|
||||
if len(records) == 0 {
|
||||
return "- 历史归档 ReAct 窗口：暂无。"
|
||||
}
|
||||
|
||||
windowLoops := tailExecuteLoops(records, executeLoopWindowLimit)
|
||||
windowLoops = compressExecuteLoopObservationsByTool(windowLoops)
|
||||
|
||||
lines := []string{"历史归档 ReAct 窗口（由上一轮 msg2 并入）："}
|
||||
for i, loop := range windowLoops {
|
||||
for i, loop := range loops {
|
||||
lines = append(lines, fmt.Sprintf("%d) thought/reason：%s", i+1, loop.Thought))
|
||||
lines = append(lines, fmt.Sprintf(" tool_call：%s", renderExecuteToolCallText(loop.ToolName, loop.ToolArgs)))
|
||||
lines = append(lines, fmt.Sprintf(" observation：%s", loop.Observation))
|
||||
@@ -525,51 +432,6 @@ func findExecuteThoughtBefore(history []*schema.Message, index int) string {
|
||||
return "（未记录）"
|
||||
}
|
||||
|
||||
func tailExecuteLoops(records []executeLoopRecord, limit int) []executeLoopRecord {
|
||||
if len(records) == 0 {
|
||||
return nil
|
||||
}
|
||||
if limit <= 0 || len(records) <= limit {
|
||||
result := make([]executeLoopRecord, len(records))
|
||||
copy(result, records)
|
||||
return result
|
||||
}
|
||||
result := make([]executeLoopRecord, limit)
|
||||
copy(result, records[len(records)-limit:])
|
||||
return result
|
||||
}
|
||||
|
||||
// compressExecuteLoopObservationsByTool 对窗口内重复工具做 observation 压缩。
|
||||
func compressExecuteLoopObservationsByTool(records []executeLoopRecord) []executeLoopRecord {
|
||||
if len(records) == 0 {
|
||||
return records
|
||||
}
|
||||
|
||||
latestIndexByTool := make(map[string]int, len(records))
|
||||
for i := len(records) - 1; i >= 0; i-- {
|
||||
key := strings.ToLower(strings.TrimSpace(records[i].ToolName))
|
||||
if key == "" {
|
||||
key = "unknown_tool"
|
||||
}
|
||||
if _, exists := latestIndexByTool[key]; !exists {
|
||||
latestIndexByTool[key] = i
|
||||
}
|
||||
}
|
||||
|
||||
result := make([]executeLoopRecord, len(records))
|
||||
copy(result, records)
|
||||
for i := range result {
|
||||
key := strings.ToLower(strings.TrimSpace(result[i].ToolName))
|
||||
if key == "" {
|
||||
key = "unknown_tool"
|
||||
}
|
||||
if latestIndexByTool[key] != i {
|
||||
result[i].Observation = executeTrimmedObservationText
|
||||
}
|
||||
}
|
||||
return result
|
||||
}
|
||||
|
||||
func renderExecuteToolCallText(toolName, toolArgs string) string {
|
||||
toolName = strings.TrimSpace(toolName)
|
||||
if toolName == "" {
|
||||
@@ -582,38 +444,6 @@ func renderExecuteToolCallText(toolName, toolArgs string) string {
|
||||
return toolName + "(" + toolArgs + ")"
|
||||
}
|
||||
|
||||
func buildEarlyExecuteReactSummary(records []executeLoopRecord, windowLimit int) string {
|
||||
if len(records) == 0 {
|
||||
return "暂无。"
|
||||
}
|
||||
if len(records) <= windowLimit {
|
||||
return "无（当前窗口已覆盖全部 ReAct 记录）。"
|
||||
}
|
||||
|
||||
early := records[:len(records)-windowLimit]
|
||||
toolCounts := make(map[string]int, len(early))
|
||||
for _, record := range early {
|
||||
key := strings.TrimSpace(record.ToolName)
|
||||
if key == "" {
|
||||
key = "unknown_tool"
|
||||
}
|
||||
toolCounts[key]++
|
||||
}
|
||||
|
||||
names := make([]string, 0, len(toolCounts))
|
||||
for name := range toolCounts {
|
||||
names = append(names, name)
|
||||
}
|
||||
sort.Strings(names)
|
||||
|
||||
parts := make([]string, 0, len(names))
|
||||
for _, name := range names {
|
||||
parts = append(parts, fmt.Sprintf("%s×%d", name, toolCounts[name]))
|
||||
}
|
||||
|
||||
return fmt.Sprintf("已折叠 %d 条旧记录，涉及：%s。", len(early), strings.Join(parts, "、"))
|
||||
}
|
||||
|
||||
func hasExecuteRoughBuildDone(ctx *newagentmodel.ConversationContext) bool {
|
||||
if ctx == nil {
|
||||
return false
|
||||
@@ -725,3 +555,12 @@ func renderExecuteTaskClassIDs(state *newagentmodel.CommonState) string {
|
||||
}
|
||||
return fmt.Sprintf("task_class_ids=[%s]", strings.Join(parts, ","))
|
||||
}
|
||||
|
||||
// renderExecuteMemoryContext 提取 execute 阶段要注入 msg3 的记忆文本。
|
||||
//
|
||||
// 1. 只读取统一的 memory_context，避免把其他 pinned block 误塞进 prompt。
|
||||
// 2. 为空时直接返回空串，保持 msg3 干净。
|
||||
// 3. 复用统一记忆渲染逻辑，保证各阶段记忆入口一致。
|
||||
func renderExecuteMemoryContext(ctx *newagentmodel.ConversationContext) string {
|
||||
return renderUnifiedMemoryContext(ctx)
|
||||
}
|
||||
|
||||
@@ -1,31 +0,0 @@
|
||||
package newagentprompt
|
||||
|
||||
import (
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
)
|
||||
|
||||
const executeMemoryContextKey = "memory_context"
|
||||
|
||||
// renderExecuteMemoryContext 提取 Execute 阶段需要补充到 msg3 的记忆文本。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. 只白名单消费 memory_context，避免把 execution_context / current_step 等 Execute 自有块再次注入；
|
||||
// 2. 若 block 不存在或正文为空，直接返回空串，不给 msg3 留空段；
|
||||
// 3. 这里不重新渲染记忆，只消费 agentsvc 已经产出的最终文本，保证所有阶段口径一致。
|
||||
func renderExecuteMemoryContext(ctx *newagentmodel.ConversationContext) string {
|
||||
if ctx == nil {
|
||||
return ""
|
||||
}
|
||||
|
||||
block, ok := ctx.PinnedBlockByKey(executeMemoryContextKey)
|
||||
if !ok {
|
||||
return ""
|
||||
}
|
||||
content := strings.TrimSpace(block.Content)
|
||||
if content == "" {
|
||||
return ""
|
||||
}
|
||||
return content
|
||||
}
|
||||
@@ -2,7 +2,6 @@ package newagentprompt
|
||||
|
||||
import (
|
||||
"fmt"
|
||||
"strconv"
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
@@ -49,14 +48,19 @@ func BuildPlanSystemPrompt() string {
|
||||
// BuildPlanMessages 组装规划阶段的 messages。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 负责把 state + context 收敛成规划阶段模型输入；
|
||||
// 2. 负责把置顶上下文和工具摘要放在 history 前面，降低模型跑偏概率；
|
||||
// 3. 不负责解析模型输出，也不负责判断规划质量。
|
||||
// 1. 负责把 state + context 收敛成统一 4 段式规划阶段模型输入；
|
||||
// 2. 不负责解析模型输出，也不负责判断规划质量；
|
||||
// 3. msg3 中的状态文本由本函数显式传入，确保统一骨架下仍能看到完整计划与阶段信息。
|
||||
func BuildPlanMessages(state *newagentmodel.CommonState, ctx *newagentmodel.ConversationContext, userInput string) []*schema.Message {
|
||||
return buildStageMessages(
|
||||
BuildPlanSystemPrompt(),
|
||||
return buildUnifiedStageMessages(
|
||||
ctx,
|
||||
BuildPlanUserPrompt(state, userInput),
|
||||
StageMessagesConfig{
|
||||
SystemPrompt: BuildPlanSystemPrompt(),
|
||||
Msg1Content: buildPlanConversationMessage(ctx),
|
||||
Msg2Content: buildPlanWorkspace(state),
|
||||
Msg3Suffix: BuildPlanUserPrompt(state, userInput),
|
||||
Msg3Role: schema.User,
|
||||
},
|
||||
)
|
||||
}
|
||||
|
||||
@@ -64,21 +68,9 @@ func BuildPlanMessages(state *newagentmodel.CommonState, ctx *newagentmodel.Conv
|
||||
func BuildPlanUserPrompt(state *newagentmodel.CommonState, userInput string) string {
|
||||
var sb strings.Builder
|
||||
|
||||
sb.WriteString("请继续当前任务的规划阶段。\n")
|
||||
sb.WriteString(renderStateSummary(state))
|
||||
sb.WriteString("\n")
|
||||
sb.WriteString("本轮目标：围绕当前任务继续规划，直到形成一份稳定、可执行的自然语言 plan，或在信息不足时明确追问用户。\n\n")
|
||||
sb.WriteString("请继续当前任务的规划阶段，严格输出 JSON。\n")
|
||||
sb.WriteString("目标：围绕最近对话和规划工作区信息，产出一份稳定、可执行的自然语言计划；若关键信息不足，请明确 ask_user。\n\n")
|
||||
sb.WriteString(BuildPlanDecisionContractText())
|
||||
sb.WriteString("\n")
|
||||
|
||||
if state != nil && len(state.TaskClassIDs) > 0 {
|
||||
parts := make([]string, len(state.TaskClassIDs))
|
||||
for i, id := range state.TaskClassIDs {
|
||||
parts[i] = strconv.Itoa(id)
|
||||
}
|
||||
sb.WriteString(fmt.Sprintf("\n本次排课请求涉及的任务类 ID（前端传入）：[%s]\n", strings.Join(parts, ", ")))
|
||||
sb.WriteString("规划时请结合上述任务类 ID 判断是否需要粗排（needs_rough_build），并在 plan_steps 中体现排课意图。\n")
|
||||
}
|
||||
|
||||
trimmedInput := strings.TrimSpace(userInput)
|
||||
if trimmedInput != "" {
|
||||
|
||||
133
backend/newAgent/prompt/plan_context.go
Normal file
133
backend/newAgent/prompt/plan_context.go
Normal file
@@ -0,0 +1,133 @@
|
||||
package newagentprompt
|
||||
|
||||
import (
|
||||
"fmt"
|
||||
"strconv"
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
)
|
||||
|
||||
// buildPlanConversationMessage 生成 plan 节点看到的真实对话视图。
|
||||
func buildPlanConversationMessage(ctx *newagentmodel.ConversationContext) string {
|
||||
return buildConversationHistoryMessage(ctx, "规划参考对话")
|
||||
}
|
||||
|
||||
// buildPlanWorkspace 渲染 plan 节点自己的工作区。
|
||||
//
|
||||
// 设计说明：
|
||||
// 1. 这里只保留“规划真正需要知道的东西”：已有计划、当前步骤、task_class_ids、任务类约束；
|
||||
// 2. 不再复用通用胖状态摘要，避免把 execute / deliver 无关状态一起塞给 plan；
|
||||
// 3. 若当前没有正式计划，则明确告诉模型“从零开始规划”，避免继续误沿用旧上下文。
|
||||
func buildPlanWorkspace(state *newagentmodel.CommonState) string {
|
||||
lines := []string{"规划工作区："}
|
||||
if state == nil {
|
||||
lines = append(lines, "- 当前缺少流程状态，请主要依据最近对话与本轮输入继续规划。")
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
if !state.HasPlan() {
|
||||
lines = append(lines, "- 当前还没有正式计划。")
|
||||
} else {
|
||||
lines = append(lines, fmt.Sprintf("- 已有计划：共 %d 步。", len(state.PlanSteps)))
|
||||
lines = append(lines, renderPlanCurrentStepSummary(state))
|
||||
lines = append(lines, "计划简表：")
|
||||
lines = append(lines, renderPlanStepOutline(state.PlanSteps))
|
||||
}
|
||||
|
||||
if taskClassIDs := renderPlanTaskClassIDs(state); taskClassIDs != "" {
|
||||
lines = append(lines, "- "+taskClassIDs)
|
||||
}
|
||||
if taskClassMeta := renderPlanTaskClassMeta(state); taskClassMeta != "" {
|
||||
lines = append(lines, "任务类约束：")
|
||||
lines = append(lines, taskClassMeta)
|
||||
}
|
||||
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// renderPlanCurrentStepSummary 返回 plan 节点需要知道的当前步骤进度。
|
||||
func renderPlanCurrentStepSummary(state *newagentmodel.CommonState) string {
|
||||
if state == nil || !state.HasPlan() {
|
||||
return "- 当前步骤：暂无。"
|
||||
}
|
||||
|
||||
current, total := state.PlanProgress()
|
||||
step, ok := state.CurrentPlanStep()
|
||||
if !ok {
|
||||
return fmt.Sprintf("- 当前步骤：计划共 %d 步，当前没有可继续沿用的有效步骤。", total)
|
||||
}
|
||||
|
||||
content := strings.TrimSpace(step.Content)
|
||||
if content == "" {
|
||||
content = "（当前步骤正文为空）"
|
||||
}
|
||||
|
||||
summary := fmt.Sprintf("- 当前步骤：第 %d/%d 步，%s", current, total, content)
|
||||
if doneWhen := strings.TrimSpace(step.DoneWhen); doneWhen != "" {
|
||||
summary += fmt.Sprintf("；完成判定：%s", doneWhen)
|
||||
}
|
||||
return summary
|
||||
}
|
||||
|
||||
// renderPlanStepOutline 将完整计划压成 plan 节点可读的简表。
|
||||
func renderPlanStepOutline(steps []newagentmodel.PlanStep) string {
|
||||
if len(steps) == 0 {
|
||||
return "- 暂无。"
|
||||
}
|
||||
|
||||
lines := make([]string, 0, len(steps))
|
||||
for i, step := range steps {
|
||||
content := strings.TrimSpace(step.Content)
|
||||
if content == "" {
|
||||
content = "（步骤正文为空）"
|
||||
}
|
||||
line := fmt.Sprintf("%d. %s", i+1, content)
|
||||
if doneWhen := strings.TrimSpace(step.DoneWhen); doneWhen != "" {
|
||||
line += fmt.Sprintf(" | 完成判定：%s", doneWhen)
|
||||
}
|
||||
lines = append(lines, line)
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
|
||||
// renderPlanTaskClassIDs 返回批量排课场景下的 task_class_ids 简表。
|
||||
func renderPlanTaskClassIDs(state *newagentmodel.CommonState) string {
|
||||
if state == nil || len(state.TaskClassIDs) == 0 {
|
||||
return ""
|
||||
}
|
||||
|
||||
parts := make([]string, len(state.TaskClassIDs))
|
||||
for i, id := range state.TaskClassIDs {
|
||||
parts[i] = strconv.Itoa(id)
|
||||
}
|
||||
return fmt.Sprintf("task_class_ids=[%s]", strings.Join(parts, ", "))
|
||||
}
|
||||
|
||||
// renderPlanTaskClassMeta 返回 plan 节点真正需要看的任务类边界。
|
||||
//
|
||||
// 说明：
|
||||
// 1. 这里只保留名称、策略、总时段、日期范围这类规划相关信息；
|
||||
// 2. 不再把所有字段原样平铺，避免工作区过胖；
|
||||
// 3. 若某项字段为空，则直接省略，不制造噪声。
|
||||
func renderPlanTaskClassMeta(state *newagentmodel.CommonState) string {
|
||||
if state == nil || len(state.TaskClasses) == 0 {
|
||||
return ""
|
||||
}
|
||||
|
||||
lines := make([]string, 0, len(state.TaskClasses))
|
||||
for _, tc := range state.TaskClasses {
|
||||
line := fmt.Sprintf("- [ID=%d] %s", tc.ID, strings.TrimSpace(tc.Name))
|
||||
if strategy := strings.TrimSpace(tc.Strategy); strategy != "" {
|
||||
line += fmt.Sprintf("；策略：%s", strategy)
|
||||
}
|
||||
if tc.TotalSlots > 0 {
|
||||
line += fmt.Sprintf("；总时段预算：%d", tc.TotalSlots)
|
||||
}
|
||||
if tc.StartDate != "" || tc.EndDate != "" {
|
||||
line += fmt.Sprintf("；日期范围：%s ~ %s", tc.StartDate, tc.EndDate)
|
||||
}
|
||||
lines = append(lines, line)
|
||||
}
|
||||
return strings.Join(lines, "\n")
|
||||
}
|
||||
212
backend/newAgent/prompt/unified_context.go
Normal file
212
backend/newAgent/prompt/unified_context.go
Normal file
@@ -0,0 +1,212 @@
|
||||
package newagentprompt
|
||||
|
||||
import (
|
||||
"strings"
|
||||
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
)
|
||||
|
||||
// ConversationTurn 表示对话历史中的一轮自然语言交互。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 这里只承载 user 与 assistant speak，不承载 tool_call 和 tool observation；
|
||||
// 2. 供 chat / plan / deliver 等节点复用，避免各节点重复写一套提取逻辑；
|
||||
// 3. 不负责裁剪长度，长度预算统一交给压缩层处理。
|
||||
type ConversationTurn struct {
|
||||
Role string
|
||||
Content string
|
||||
}
|
||||
|
||||
// StageMessagesConfig 描述统一四段式骨架下，各节点自行提供的内容块。
|
||||
//
|
||||
// 设计目标：
|
||||
// 1. 统一层只负责“四条消息怎么拼”，不再替节点决定“每条消息里该放什么”；
|
||||
// 2. Msg1 / Msg2 / Msg3Prefix / Msg3Suffix 都由节点自己渲染，避免 chat / plan / deliver 继续套 execute 的内容模板；
|
||||
// 3. memory_context 仍由统一层单入口注入到 msg3，避免多处重复注入。
|
||||
type StageMessagesConfig struct {
|
||||
// SystemPrompt 是节点自己的系统提示词。
|
||||
SystemPrompt string
|
||||
|
||||
// Msg1Content 是第 2 条 assistant 消息，通常放“节点想看的历史视图”。
|
||||
Msg1Content string
|
||||
|
||||
// Msg2Content 是第 3 条 assistant 消息，通常放“节点自己的工作区/补充约束”。
|
||||
Msg2Content string
|
||||
|
||||
// Msg3Prefix 是第 4 条消息中位于 memory_context 之前的内容。
|
||||
// 常见放法：阶段状态、规划工作区摘要、交付收口约束等。
|
||||
Msg3Prefix string
|
||||
|
||||
// Msg3Suffix 是第 4 条消息中位于 memory_context 之后的内容。
|
||||
// 对 user-role 节点来说，这里通常放最终用户指令，保证“用户输入收尾”。
|
||||
Msg3Suffix string
|
||||
|
||||
// Msg3Role 指定第 4 条消息的角色。
|
||||
// Execute 继续使用 system，其余节点一般使用 user。
|
||||
Msg3Role schema.RoleType
|
||||
}
|
||||
|
||||
// buildUnifiedStageMessages 组装统一 4 段式消息骨架。
|
||||
//
|
||||
// 固定布局：
|
||||
// 1. msg0(system)：系统规则 + 阶段规则 + 工具简表；
|
||||
// 2. msg1(assistant)：节点自定义的历史视图；
|
||||
// 3. msg2(assistant)：节点自定义的工作区；
|
||||
// 4. msg3(user/system)：节点自定义前后缀 + 统一 memory_context。
|
||||
func buildUnifiedStageMessages(
|
||||
ctx *newagentmodel.ConversationContext,
|
||||
config StageMessagesConfig,
|
||||
) []*schema.Message {
|
||||
msg0 := buildUnifiedMsg0(config.SystemPrompt, ctx)
|
||||
msg1 := buildUnifiedMsg1(config.Msg1Content)
|
||||
msg2 := buildUnifiedMsg2(config.Msg2Content)
|
||||
msg3 := buildUnifiedMsg3(ctx, config)
|
||||
|
||||
return []*schema.Message{
|
||||
schema.SystemMessage(msg0),
|
||||
{Role: schema.Assistant, Content: msg1},
|
||||
{Role: schema.Assistant, Content: msg2},
|
||||
buildUnifiedMsg3Message(msg3, config.Msg3Role),
|
||||
}
|
||||
}
|
||||
|
||||
// buildUnifiedMsg3Message 根据配置决定第 4 条消息的角色。
|
||||
func buildUnifiedMsg3Message(content string, role schema.RoleType) *schema.Message {
|
||||
if role == schema.User {
|
||||
return schema.UserMessage(content)
|
||||
}
|
||||
return schema.SystemMessage(content)
|
||||
}
|
||||
|
||||
// buildUnifiedMsg0 合并系统提示 + 工具简表，生成 msg0。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. 先合并基础系统提示与节点系统提示，保证模型身份稳定；
|
||||
// 2. 若当前节点注入了工具 schema，则附加紧凑工具目录；
|
||||
// 3. 若两部分都为空，则回退到最小兜底提示，避免出现空消息。
|
||||
func buildUnifiedMsg0(stageSystemPrompt string, ctx *newagentmodel.ConversationContext) string {
|
||||
base := strings.TrimSpace(mergeSystemPrompts(ctx, stageSystemPrompt))
|
||||
if base == "" {
|
||||
base = "你是 SmartMate 助手，请继续当前阶段。"
|
||||
}
|
||||
|
||||
toolCatalog := renderExecuteToolCatalogCompact(ctx)
|
||||
if toolCatalog == "" {
|
||||
return base
|
||||
}
|
||||
return base + "\n\n" + toolCatalog
|
||||
}
|
||||
|
||||
// buildUnifiedMsg1 返回节点自行提供的历史视图。
|
||||
//
|
||||
// 说明：
|
||||
// 1. 统一层不再内置 execute 风格的 ReAct 摘要；
|
||||
// 2. 节点若未传入内容，则回退到最小占位，保证四段结构稳定；
|
||||
// 3. 压缩层仍会统一统计和压缩这条消息。
|
||||
func buildUnifiedMsg1(content string) string {
|
||||
content = strings.TrimSpace(content)
|
||||
if content != "" {
|
||||
return content
|
||||
}
|
||||
return "历史上下文：暂无。"
|
||||
}
|
||||
|
||||
// buildUnifiedMsg2 返回节点自行提供的工作区。
|
||||
//
|
||||
// 说明：
|
||||
// 1. 非 execute 节点也允许有自己的 msg2，不再被统一层硬塞“暂无”语义；
|
||||
// 2. 若节点暂时没有额外工作区，则回退到最小占位，保证结构稳定。
|
||||
func buildUnifiedMsg2(content string) string {
|
||||
content = strings.TrimSpace(content)
|
||||
if content != "" {
|
||||
return content
|
||||
}
|
||||
return "阶段工作区：暂无。"
|
||||
}
|
||||
|
||||
// buildUnifiedMsg3 统一拼装 msg3：前缀 + memory_context + 后缀。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. 前缀由节点决定，适合放轻量状态或阶段约束；
|
||||
// 2. memory_context 只在这里注入一次，避免 pinned block 多入口重复出现；
|
||||
// 3. 后缀由节点决定。对于 user-role 节点，通常把最终用户指令放在这里，保证消息末尾仍是用户输入。
|
||||
func buildUnifiedMsg3(ctx *newagentmodel.ConversationContext, config StageMessagesConfig) string {
|
||||
var sections []string
|
||||
|
||||
if prefix := strings.TrimSpace(config.Msg3Prefix); prefix != "" {
|
||||
sections = append(sections, prefix)
|
||||
}
|
||||
if memoryText := renderUnifiedMemoryContext(ctx); memoryText != "" {
|
||||
sections = append(sections, "相关记忆（仅在确有帮助时参考，不要机械复述）：\n"+memoryText)
|
||||
}
|
||||
if suffix := strings.TrimSpace(config.Msg3Suffix); suffix != "" {
|
||||
sections = append(sections, suffix)
|
||||
}
|
||||
|
||||
if len(sections) == 0 {
|
||||
return "请继续当前阶段。"
|
||||
}
|
||||
return strings.Join(sections, "\n\n")
|
||||
}
|
||||
|
||||
// renderUnifiedMemoryContext 提取需要补充到 msg3 的记忆文本。
|
||||
//
|
||||
// 步骤化说明：
|
||||
// 1. 只消费 memory_context，避免把 execution_context / current_step 等阶段专属块混回 prompt；
|
||||
// 2. block 不存在或正文为空时直接返回空串；
|
||||
// 3. 这里只读取 agentsvc 已经产出的最终文本，不在这里重新拼装记忆。
|
||||
func renderUnifiedMemoryContext(ctx *newagentmodel.ConversationContext) string {
|
||||
if ctx == nil {
|
||||
return ""
|
||||
}
|
||||
|
||||
block, ok := ctx.PinnedBlockByKey("memory_context")
|
||||
if !ok {
|
||||
return ""
|
||||
}
|
||||
content := strings.TrimSpace(block.Content)
|
||||
if content == "" {
|
||||
return ""
|
||||
}
|
||||
return content
|
||||
}
|
||||
|
||||
// CollectConversationTurns 从历史消息中提取 user + assistant speak 对话流。
|
||||
//
|
||||
// 提取规则：
|
||||
// 1. 只保留 user 消息（排除 correction prompt）和 assistant 纯文本消息；
|
||||
// 2. assistant tool_call 消息与 tool observation 消息不纳入“真实对话”；
|
||||
// 3. 返回顺序保持与原始 history 一致。
|
||||
func CollectConversationTurns(history []*schema.Message) []ConversationTurn {
|
||||
if len(history) == 0 {
|
||||
return nil
|
||||
}
|
||||
|
||||
turns := make([]ConversationTurn, 0, len(history))
|
||||
for _, msg := range history {
|
||||
if msg == nil {
|
||||
continue
|
||||
}
|
||||
text := strings.TrimSpace(msg.Content)
|
||||
if text == "" {
|
||||
continue
|
||||
}
|
||||
switch msg.Role {
|
||||
case schema.User:
|
||||
// 1. 跳过后端注入的 correction prompt，避免把纠错文案误判为用户真实意图。
|
||||
if isExecuteCorrectionPrompt(msg) {
|
||||
continue
|
||||
}
|
||||
turns = append(turns, ConversationTurn{Role: "user", Content: text})
|
||||
case schema.Assistant:
|
||||
// 2. 跳过工具调用消息，只保留真正面向用户的 speak/答复。
|
||||
if len(msg.ToolCalls) > 0 {
|
||||
continue
|
||||
}
|
||||
turns = append(turns, ConversationTurn{Role: "assistant", Content: text})
|
||||
}
|
||||
}
|
||||
|
||||
return turns
|
||||
}
|
||||
@@ -27,6 +27,8 @@ const (
|
||||
ExecuteTokenBudget = 80000
|
||||
// msg0 + msg3 固定开销 + 安全余量
|
||||
ExecuteReserveTokens = 8000
|
||||
StageTokenBudget = ExecuteTokenBudget
|
||||
StageReserveTokens = ExecuteReserveTokens
|
||||
)
|
||||
|
||||
// MaxContextTokensByModel 返回指定模型的最大上下文 token。
|
||||
@@ -151,8 +153,8 @@ func isCJK(r rune) bool {
|
||||
return unicode.Is(unicode.Han, r) || unicode.Is(unicode.Hiragana, r) || unicode.Is(unicode.Katakana, r) || unicode.Is(unicode.Hangul, r)
|
||||
}
|
||||
|
||||
// ExecuteTokenBreakdown 是 Execute 阶段四条消息的 token 分布。
|
||||
type ExecuteTokenBreakdown struct {
|
||||
// StageTokenBreakdown 记录四条阶段消息的 token 分布。
|
||||
type StageTokenBreakdown struct {
|
||||
Msg0 int `json:"msg0"`
|
||||
Msg1 int `json:"msg1"`
|
||||
Msg2 int `json:"msg2"`
|
||||
@@ -161,31 +163,47 @@ type ExecuteTokenBreakdown struct {
|
||||
Budget int `json:"budget"`
|
||||
}
|
||||
|
||||
// EstimateExecuteMessagesTokens 估算 Execute 四条消息的 token 分布。
|
||||
func EstimateExecuteMessagesTokens(msg0, msg1, msg2, msg3 string) ExecuteTokenBreakdown {
|
||||
b := ExecuteTokenBreakdown{
|
||||
// ExecuteTokenBreakdown 保留为历史兼容别名，避免旧调用点改动。
|
||||
type ExecuteTokenBreakdown = StageTokenBreakdown
|
||||
|
||||
// EstimateStageMessagesTokens 估算四条阶段消息的 token 分布。
|
||||
func EstimateStageMessagesTokens(msg0, msg1, msg2, msg3 string) StageTokenBreakdown {
|
||||
b := StageTokenBreakdown{
|
||||
Msg0: EstimateTextTokens(msg0),
|
||||
Msg1: EstimateTextTokens(msg1),
|
||||
Msg2: EstimateTextTokens(msg2),
|
||||
Msg3: EstimateTextTokens(msg3),
|
||||
Budget: ExecuteTokenBudget,
|
||||
Budget: StageTokenBudget,
|
||||
}
|
||||
b.Total = b.Msg0 + b.Msg1 + b.Msg2 + b.Msg3
|
||||
return b
|
||||
}
|
||||
|
||||
// CheckExecuteTokenBudget 检查是否超出 token 预算。
|
||||
// 返回 breakdown、是否超限、是否需要压缩 msg1、是否需要压缩 msg2。
|
||||
func CheckExecuteTokenBudget(msg0, msg1, msg2, msg3 string) (breakdown ExecuteTokenBreakdown, overBudget bool, needCompactMsg1 bool, needCompactMsg2 bool) {
|
||||
breakdown = EstimateExecuteMessagesTokens(msg0, msg1, msg2, msg3)
|
||||
overBudget = breakdown.Total > ExecuteTokenBudget
|
||||
// CheckStageTokenBudget 检查是否超出阶段预算，并给出需要压缩的消息标记。
|
||||
//
|
||||
// 1. 先计算四条消息的 token 分布，便于后续日志和统计。
|
||||
// 2. 如果总量没有超预算，直接返回。
|
||||
// 3. 如果超预算，则按 msg1 / msg2 的相对占比判断是否需要分别压缩。
|
||||
func CheckStageTokenBudget(msg0, msg1, msg2, msg3 string) (breakdown StageTokenBreakdown, overBudget bool, needCompactMsg1 bool, needCompactMsg2 bool) {
|
||||
breakdown = EstimateStageMessagesTokens(msg0, msg1, msg2, msg3)
|
||||
overBudget = breakdown.Total > StageTokenBudget
|
||||
if !overBudget {
|
||||
return
|
||||
}
|
||||
// msg1 超过可用预算的一半时需要压缩
|
||||
available := ExecuteTokenBudget - ExecuteReserveTokens
|
||||
// msg1 过大时，优先压缩历史对话。
|
||||
available := StageTokenBudget - StageReserveTokens
|
||||
needCompactMsg1 = breakdown.Msg1 > available/2
|
||||
// 压缩 msg1 后仍超限，则压缩 msg2
|
||||
needCompactMsg2 = (breakdown.Total - breakdown.Msg1 + available/4) > ExecuteTokenBudget
|
||||
// 若压缩 msg1 后仍然超限，再压缩执行记录区。
|
||||
needCompactMsg2 = (breakdown.Total - breakdown.Msg1 + available/4) > StageTokenBudget
|
||||
return
|
||||
}
|
||||
|
||||
// EstimateExecuteMessagesTokens 保留旧名称，内部复用阶段预算实现。
|
||||
func EstimateExecuteMessagesTokens(msg0, msg1, msg2, msg3 string) StageTokenBreakdown {
|
||||
return EstimateStageMessagesTokens(msg0, msg1, msg2, msg3)
|
||||
}
|
||||
|
||||
// CheckExecuteTokenBudget 保留旧名称，内部复用阶段预算实现。
|
||||
func CheckExecuteTokenBudget(msg0, msg1, msg2, msg3 string) (breakdown StageTokenBreakdown, overBudget bool, needCompactMsg1 bool, needCompactMsg2 bool) {
|
||||
return CheckStageTokenBudget(msg0, msg1, msg2, msg3)
|
||||
}
|
||||
|
||||
@@ -3,14 +3,12 @@ package agentsvc
|
||||
import (
|
||||
"context"
|
||||
"encoding/json"
|
||||
"errors"
|
||||
"log"
|
||||
"strconv"
|
||||
"strings"
|
||||
"time"
|
||||
|
||||
agentchat "github.com/LoveLosita/smartflow/backend/agent/chat"
|
||||
agentrouter "github.com/LoveLosita/smartflow/backend/agent/router"
|
||||
"github.com/LoveLosita/smartflow/backend/conv"
|
||||
"github.com/LoveLosita/smartflow/backend/dao"
|
||||
outboxinfra "github.com/LoveLosita/smartflow/backend/infra/outbox"
|
||||
@@ -21,7 +19,6 @@ import (
|
||||
newagentmodel "github.com/LoveLosita/smartflow/backend/newAgent/model"
|
||||
newagenttools "github.com/LoveLosita/smartflow/backend/newAgent/tools"
|
||||
"github.com/LoveLosita/smartflow/backend/pkg"
|
||||
"github.com/LoveLosita/smartflow/backend/respond"
|
||||
eventsvc "github.com/LoveLosita/smartflow/backend/service/events"
|
||||
"github.com/cloudwego/eino-ext/components/model/ark"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
@@ -129,10 +126,6 @@ func (s *AgentService) PersistChatHistory(ctx context.Context, payload model.Cha
|
||||
payload.Message,
|
||||
payload.ReasoningContent,
|
||||
payload.ReasoningDurationSeconds,
|
||||
payload.RetryGroupID,
|
||||
payload.RetryIndex,
|
||||
payload.RetryFromUserMessageID,
|
||||
payload.RetryFromAssistantMessageID,
|
||||
payload.TokensConsumed,
|
||||
)
|
||||
}
|
||||
@@ -159,112 +152,6 @@ func mergeAgentReasoningText(parts ...string) string {
|
||||
return strings.Join(merged, "\n\n")
|
||||
}
|
||||
|
||||
type chatRetryMeta struct {
|
||||
GroupID string
|
||||
Index int
|
||||
FromUserMessageID int
|
||||
FromAssistantMessageID int
|
||||
}
|
||||
|
||||
func (m *chatRetryMeta) GroupIDPtr() *string {
|
||||
if m == nil || strings.TrimSpace(m.GroupID) == "" {
|
||||
return nil
|
||||
}
|
||||
groupID := strings.TrimSpace(m.GroupID)
|
||||
return &groupID
|
||||
}
|
||||
|
||||
func (m *chatRetryMeta) IndexPtr() *int {
|
||||
if m == nil || m.Index <= 0 {
|
||||
return nil
|
||||
}
|
||||
index := m.Index
|
||||
return &index
|
||||
}
|
||||
|
||||
func (m *chatRetryMeta) FromUserMessageIDPtr() *int {
|
||||
if m == nil || m.FromUserMessageID <= 0 {
|
||||
return nil
|
||||
}
|
||||
id := m.FromUserMessageID
|
||||
return &id
|
||||
}
|
||||
|
||||
func (m *chatRetryMeta) FromAssistantMessageIDPtr() *int {
|
||||
if m == nil || m.FromAssistantMessageID <= 0 {
|
||||
return nil
|
||||
}
|
||||
id := m.FromAssistantMessageID
|
||||
return &id
|
||||
}
|
||||
|
||||
func (m *chatRetryMeta) CacheExtra() map[string]any {
|
||||
if m == nil || strings.TrimSpace(m.GroupID) == "" || m.Index <= 0 {
|
||||
return nil
|
||||
}
|
||||
extra := map[string]any{
|
||||
"retry_group_id": m.GroupID,
|
||||
"retry_index": m.Index,
|
||||
}
|
||||
if m.FromUserMessageID > 0 {
|
||||
extra["retry_from_user_message_id"] = m.FromUserMessageID
|
||||
}
|
||||
if m.FromAssistantMessageID > 0 {
|
||||
extra["retry_from_assistant_message_id"] = m.FromAssistantMessageID
|
||||
}
|
||||
return extra
|
||||
}
|
||||
|
||||
func (s *AgentService) buildChatRetryMeta(ctx context.Context, userID int, chatID string, extra map[string]any) (*chatRetryMeta, error) {
|
||||
if len(extra) == 0 {
|
||||
return nil, nil
|
||||
}
|
||||
requestMode := strings.ToLower(strings.TrimSpace(readAgentExtraString(extra, "request_mode")))
|
||||
if requestMode != "retry" {
|
||||
return nil, nil
|
||||
}
|
||||
|
||||
groupID := strings.TrimSpace(readAgentExtraString(extra, "retry_group_id"))
|
||||
if groupID == "" {
|
||||
groupID = uuid.NewString()
|
||||
}
|
||||
|
||||
sourceUserMessageID := readAgentExtraInt(extra, "retry_from_user_message_id")
|
||||
sourceAssistantMessageID := readAgentExtraInt(extra, "retry_from_assistant_message_id")
|
||||
// 1. retry 请求必须明确指向“被重试的那一轮 user + assistant”。
|
||||
// 2. 若这里拿不到有效父消息 id，继续写库只会生成一组孤立的 index=1 重试消息。
|
||||
// 3. 因此直接拒绝本次请求，让前端刷新历史后重试，比静默写脏数据更安全。
|
||||
if sourceUserMessageID <= 0 || sourceAssistantMessageID <= 0 {
|
||||
return nil, errors.New("重试请求缺少有效的父消息ID，请刷新会话后重试")
|
||||
}
|
||||
// 4. 再进一步校验父消息确实属于当前用户与当前会话，且角色语义正确。
|
||||
// 5. 这样即便前端误把占位 id 或串号 id 发过来，后端也不会继续落错库。
|
||||
if err := s.repo.ValidateRetrySourceMessages(ctx, userID, chatID, sourceUserMessageID, sourceAssistantMessageID); err != nil {
|
||||
return nil, errors.New("重试引用的父消息无效，请刷新会话后重试")
|
||||
}
|
||||
|
||||
if err := s.repo.EnsureRetryGroupSeed(ctx, userID, chatID, groupID, sourceUserMessageID, sourceAssistantMessageID); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
if s.agentCache != nil && (sourceUserMessageID > 0 || sourceAssistantMessageID > 0) {
|
||||
if cacheErr := s.agentCache.ApplyRetrySeed(ctx, chatID, groupID, sourceUserMessageID, sourceAssistantMessageID); cacheErr != nil {
|
||||
log.Printf("更新重试分组缓存失败 chat=%s group=%s err=%v", chatID, groupID, cacheErr)
|
||||
}
|
||||
}
|
||||
|
||||
nextIndex, err := s.repo.GetRetryGroupNextIndex(ctx, userID, chatID, groupID)
|
||||
if err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
return &chatRetryMeta{
|
||||
GroupID: groupID,
|
||||
Index: nextIndex,
|
||||
FromUserMessageID: sourceUserMessageID,
|
||||
FromAssistantMessageID: sourceAssistantMessageID,
|
||||
}, nil
|
||||
}
|
||||
|
||||
func readAgentExtraString(extra map[string]any, key string) string {
|
||||
if len(extra) == 0 {
|
||||
return ""
|
||||
@@ -400,9 +287,9 @@ func (s *AgentService) runNormalChatFlow(
|
||||
selectedModel *ark.ChatModel,
|
||||
resolvedModelName string,
|
||||
userMessage string,
|
||||
userPersisted bool,
|
||||
assistantReasoningPrefix string,
|
||||
assistantReasoningStartedAt *time.Time,
|
||||
retryMeta *chatRetryMeta,
|
||||
ifThinking bool,
|
||||
userID int,
|
||||
chatID string,
|
||||
@@ -481,44 +368,38 @@ func (s *AgentService) runNormalChatFlow(
|
||||
// 7. 后置持久化（用户消息）：
|
||||
// 7.1 先写 Redis，保证“最新会话上下文”可立即用于下一轮推理；
|
||||
// 7.2 再走可靠持久化入口（outbox 或同步 DB）。
|
||||
userMsg := &schema.Message{Role: schema.User, Content: userMessage}
|
||||
if retryExtra := retryMeta.CacheExtra(); len(retryExtra) > 0 {
|
||||
userMsg.Extra = retryExtra
|
||||
}
|
||||
if err = s.agentCache.PushMessage(ctx, chatID, userMsg); err != nil {
|
||||
log.Printf("写入用户消息到 Redis 失败: %v", err)
|
||||
}
|
||||
if !userPersisted {
|
||||
userMsg := &schema.Message{Role: schema.User, Content: userMessage}
|
||||
if err = s.agentCache.PushMessage(ctx, chatID, userMsg); err != nil {
|
||||
log.Printf("写入用户消息到 Redis 失败: %v", err)
|
||||
}
|
||||
|
||||
if err = s.PersistChatHistory(ctx, model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "user",
|
||||
Message: userMessage,
|
||||
ReasoningContent: "",
|
||||
ReasoningDurationSeconds: 0,
|
||||
RetryGroupID: retryMeta.GroupIDPtr(),
|
||||
RetryIndex: retryMeta.IndexPtr(),
|
||||
RetryFromUserMessageID: retryMeta.FromUserMessageIDPtr(),
|
||||
RetryFromAssistantMessageID: retryMeta.FromAssistantMessageIDPtr(),
|
||||
// 口径B：用户消息固定记 0；本轮总 token 统一记在助手消息。
|
||||
TokensConsumed: 0,
|
||||
}); err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
return
|
||||
if err = s.PersistChatHistory(ctx, model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "user",
|
||||
Message: userMessage,
|
||||
ReasoningContent: "",
|
||||
ReasoningDurationSeconds: 0,
|
||||
// 口径 B：用户消息固定记 0；本轮总 token 统一记在助手消息。
|
||||
TokensConsumed: 0,
|
||||
}); err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
return
|
||||
}
|
||||
s.appendConversationHistoryCacheOptimistically(
|
||||
context.Background(),
|
||||
userID,
|
||||
chatID,
|
||||
buildOptimisticConversationHistoryItem(
|
||||
"user",
|
||||
userMessage,
|
||||
"",
|
||||
0,
|
||||
requestStart,
|
||||
),
|
||||
)
|
||||
}
|
||||
s.appendConversationHistoryCacheOptimistically(
|
||||
context.Background(),
|
||||
userID,
|
||||
chatID,
|
||||
buildOptimisticConversationHistoryItem(
|
||||
"user",
|
||||
userMessage,
|
||||
"",
|
||||
0,
|
||||
retryMeta,
|
||||
requestStart,
|
||||
),
|
||||
)
|
||||
|
||||
// 普通聊天链路也需要把助手回复写入 Redis，
|
||||
// 否则会出现“数据库有助手消息，但 Redis 最新会话只有用户消息”的口径不一致。
|
||||
@@ -529,29 +410,17 @@ func (s *AgentService) runNormalChatFlow(
|
||||
if reasoningDurationSeconds > 0 {
|
||||
assistantMsg.Extra = map[string]any{"reasoning_duration_seconds": reasoningDurationSeconds}
|
||||
}
|
||||
if retryExtra := retryMeta.CacheExtra(); len(retryExtra) > 0 {
|
||||
if assistantMsg.Extra == nil {
|
||||
assistantMsg.Extra = make(map[string]any, len(retryExtra))
|
||||
}
|
||||
for key, value := range retryExtra {
|
||||
assistantMsg.Extra[key] = value
|
||||
}
|
||||
}
|
||||
if err = s.agentCache.PushMessage(context.Background(), chatID, assistantMsg); err != nil {
|
||||
log.Printf("写入助手消息到 Redis 失败: %v", err)
|
||||
}
|
||||
|
||||
if saveErr := s.PersistChatHistory(context.Background(), model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "assistant",
|
||||
Message: fullText,
|
||||
ReasoningContent: assistantReasoning,
|
||||
ReasoningDurationSeconds: reasoningDurationSeconds,
|
||||
RetryGroupID: retryMeta.GroupIDPtr(),
|
||||
RetryIndex: retryMeta.IndexPtr(),
|
||||
RetryFromUserMessageID: retryMeta.FromUserMessageIDPtr(),
|
||||
RetryFromAssistantMessageID: retryMeta.FromAssistantMessageIDPtr(),
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "assistant",
|
||||
Message: fullText,
|
||||
ReasoningContent: assistantReasoning,
|
||||
ReasoningDurationSeconds: reasoningDurationSeconds,
|
||||
// 口径B：助手消息记录“本轮请求总 token”。
|
||||
TokensConsumed: requestTotalTokens,
|
||||
}); saveErr != nil {
|
||||
@@ -566,7 +435,6 @@ func (s *AgentService) runNormalChatFlow(
|
||||
fullText,
|
||||
assistantReasoning,
|
||||
reasoningDurationSeconds,
|
||||
retryMeta,
|
||||
time.Now(),
|
||||
),
|
||||
)
|
||||
@@ -591,196 +459,3 @@ func (s *AgentService) AgentChat(ctx context.Context, userMessage string, thinki
|
||||
|
||||
return outChan, errChan
|
||||
}
|
||||
|
||||
// agentChatOld 是旧路由逻辑的备份，暂时保留供回滚使用。
|
||||
// TODO: 新 graph 稳定后删除。
|
||||
func (s *AgentService) agentChatOld(ctx context.Context, userMessage string, thinkingMode string, modelName string, userID int, chatID string, extra map[string]any) (<-chan string, <-chan error) {
|
||||
ifThinking := thinkingModeToBool(thinkingMode)
|
||||
requestStart := time.Now()
|
||||
traceID := uuid.NewString()
|
||||
|
||||
outChan := make(chan string, 256)
|
||||
errChan := make(chan error, 1)
|
||||
|
||||
// 0. 初始化”请求级 token 统计器”，用于聚合本次请求所有模型开销。
|
||||
requestCtx, _ := withRequestTokenMeter(ctx)
|
||||
|
||||
// 1) 规范会话 ID，选择模型。
|
||||
chatID = normalizeConversationID(chatID)
|
||||
selectedModel, resolvedModelName := s.pickChatModel(modelName)
|
||||
|
||||
// 2) 确保会话存在（优先缓存，必要时回源 DB 并创建）。
|
||||
// 2.1 先查 Redis 会话标记，命中则可跳过 DB 存在性校验。
|
||||
result, err := s.agentCache.GetConversationStatus(requestCtx, chatID)
|
||||
if err != nil {
|
||||
errChan <- err
|
||||
close(outChan)
|
||||
close(errChan)
|
||||
return outChan, errChan
|
||||
}
|
||||
if !result {
|
||||
// 2.2 缓存未命中时回源 DB：确认会话是否存在。
|
||||
innerResult, ifErr := s.repo.IfChatExists(requestCtx, userID, chatID)
|
||||
if ifErr != nil {
|
||||
errChan <- ifErr
|
||||
close(outChan)
|
||||
close(errChan)
|
||||
return outChan, errChan
|
||||
}
|
||||
if !innerResult {
|
||||
// 2.3 DB 里也不存在则创建新会话。
|
||||
if _, err = s.repo.CreateNewChat(userID, chatID); err != nil {
|
||||
errChan <- err
|
||||
close(outChan)
|
||||
close(errChan)
|
||||
return outChan, errChan
|
||||
}
|
||||
}
|
||||
// 2.4 补写 Redis 会话标记，优化下次访问。
|
||||
if err = s.agentCache.SetConversationStatus(requestCtx, chatID); err != nil {
|
||||
log.Printf("设置会话状态缓存失败 chat=%s: %v", chatID, err)
|
||||
}
|
||||
}
|
||||
|
||||
retryMeta, err := s.buildChatRetryMeta(requestCtx, userID, chatID, extra)
|
||||
if err != nil {
|
||||
errChan <- err
|
||||
close(outChan)
|
||||
close(errChan)
|
||||
return outChan, errChan
|
||||
}
|
||||
|
||||
// 3) 统一异步分流：
|
||||
// 3.1 先走“通用控制码路由”决定 action（chat / quick_note_create / task_query）；
|
||||
// 3.2 quick_note_create 进入随口记 graph；
|
||||
// 3.3 task_query 进入任务查询 tool-calling；
|
||||
// 3.4 chat 直接普通流式聊天。
|
||||
go func() {
|
||||
defer close(outChan)
|
||||
|
||||
// 3.1 先走轻量路由，拿到统一 action。
|
||||
routing := s.decideActionRouting(requestCtx, selectedModel, userMessage)
|
||||
if routing.RouteFailed {
|
||||
// 3.1.1 路由码失败不再回落聊天。
|
||||
// 3.1.2 直接返回内部错误，避免误进入业务分支导致“吐错内容”（例如吐排程 JSON）。
|
||||
pushErrNonBlocking(errChan, respond.RouteControlInternalError)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.2 chat：直接走普通聊天主链路。
|
||||
if routing.Action == agentrouter.ActionChat {
|
||||
s.runNormalChatFlow(requestCtx, selectedModel, resolvedModelName, userMessage, "", nil, retryMeta, ifThinking, userID, chatID, traceID, requestStart, outChan, errChan)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.3 非 chat 分支统一先发“接收成功”阶段，减少用户等待时的“无反馈感”。
|
||||
progress := newQuickNoteProgressEmitter(outChan, resolvedModelName, true)
|
||||
progress.Emit("request.accepted", routing.Detail)
|
||||
|
||||
// 3.4 quick_note_create：执行随口记 graph。
|
||||
if routing.Action == agentrouter.ActionQuickNoteCreate {
|
||||
quickHandled, quickState, quickErr := s.tryHandleQuickNoteWithGraph(
|
||||
requestCtx,
|
||||
selectedModel,
|
||||
userMessage,
|
||||
userID,
|
||||
chatID,
|
||||
traceID,
|
||||
routing.TrustRoute,
|
||||
progress.Emit,
|
||||
)
|
||||
if quickErr != nil {
|
||||
// graph 出错不直接中断用户请求，而是回退普通聊天，保证可用性优先。
|
||||
log.Printf("随口记 graph 执行失败，回退普通聊天 trace_id=%s chat_id=%s err=%v", traceID, chatID, quickErr)
|
||||
}
|
||||
|
||||
if quickHandled {
|
||||
// 3.4.1 随口记处理成功：组织最终回复并按 OpenAI 兼容格式输出。
|
||||
progress.Emit("quick_note.reply.polishing", "正在结合你的话题润色回复。")
|
||||
quickReply := buildQuickNoteFinalReply(requestCtx, selectedModel, userMessage, quickState)
|
||||
if emitErr := emitSingleAssistantCompletion(outChan, resolvedModelName, quickReply); emitErr != nil {
|
||||
pushErrNonBlocking(errChan, emitErr)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.4.2 对随口记回复执行统一后置持久化（Redis + outbox/DB）。
|
||||
requestTotalTokens := snapshotRequestTokenMeter(requestCtx).TotalTokens
|
||||
s.persistChatAfterReply(requestCtx, userID, chatID, userMessage, quickReply, progress.HistoryText(), progress.DurationSeconds(time.Now()), retryMeta, 0, requestTotalTokens, errChan)
|
||||
// 3.4.3 随口记链路同样异步生成会话标题（仅首次写入）。
|
||||
s.ensureConversationTitleAsync(userID, chatID)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.4.4 路由误判或 graph 判定非随口记时，回落普通聊天，保证“能聊”。
|
||||
progress.Emit("quick_note.fallback", "当前输入不是随口记请求，切换到普通对话。")
|
||||
s.runNormalChatFlow(requestCtx, selectedModel, resolvedModelName, userMessage, progress.HistoryText(), progress.StartedAt(), retryMeta, ifThinking, userID, chatID, traceID, requestStart, outChan, errChan)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.5 task_query：执行任务查询 tool-calling。
|
||||
if routing.Action == agentrouter.ActionTaskQuery {
|
||||
reply, queryErr := s.runTaskQueryFlow(requestCtx, selectedModel, userMessage, userID, progress.Emit)
|
||||
if queryErr != nil {
|
||||
// 3.5.1 任务查询失败时回退普通聊天，避免请求直接中断。
|
||||
log.Printf("任务查询 tool-calling 执行失败，回退普通聊天 trace_id=%s chat_id=%s err=%v", traceID, chatID, queryErr)
|
||||
progress.Emit("task_query.fallback", "任务查询暂不可用，先切回普通对话。")
|
||||
s.runNormalChatFlow(requestCtx, selectedModel, resolvedModelName, userMessage, progress.HistoryText(), progress.StartedAt(), retryMeta, ifThinking, userID, chatID, traceID, requestStart, outChan, errChan)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.5.2 查询成功后按 OpenAI 兼容格式输出，并执行统一后置持久化。
|
||||
if emitErr := emitSingleAssistantCompletion(outChan, resolvedModelName, reply); emitErr != nil {
|
||||
pushErrNonBlocking(errChan, emitErr)
|
||||
return
|
||||
}
|
||||
requestTotalTokens := snapshotRequestTokenMeter(requestCtx).TotalTokens
|
||||
s.persistChatAfterReply(requestCtx, userID, chatID, userMessage, reply, progress.HistoryText(), progress.DurationSeconds(time.Now()), retryMeta, 0, requestTotalTokens, errChan)
|
||||
s.ensureConversationTitleAsync(userID, chatID)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.6 schedule_plan：执行智能排程 graph。
|
||||
if routing.Action == agentrouter.ActionSchedulePlanCreate {
|
||||
reply, planErr := s.runSchedulePlanFlow(requestCtx, selectedModel, userMessage, userID, chatID, traceID, extra, progress.Emit, outChan, resolvedModelName)
|
||||
if planErr != nil {
|
||||
log.Printf("智能排程 graph 执行失败，回退普通聊天 trace_id=%s chat_id=%s err=%v", traceID, chatID, planErr)
|
||||
progress.Emit("schedule_plan.fallback", "智能排程暂不可用，先切回普通对话。")
|
||||
s.runNormalChatFlow(requestCtx, selectedModel, resolvedModelName, userMessage, progress.HistoryText(), progress.StartedAt(), retryMeta, ifThinking, userID, chatID, traceID, requestStart, outChan, errChan)
|
||||
return
|
||||
}
|
||||
|
||||
if emitErr := emitSingleAssistantCompletion(outChan, resolvedModelName, reply); emitErr != nil {
|
||||
pushErrNonBlocking(errChan, emitErr)
|
||||
return
|
||||
}
|
||||
requestTotalTokens := snapshotRequestTokenMeter(requestCtx).TotalTokens
|
||||
s.persistChatAfterReply(requestCtx, userID, chatID, userMessage, reply, progress.HistoryText(), progress.DurationSeconds(time.Now()), retryMeta, 0, requestTotalTokens, errChan)
|
||||
s.ensureConversationTitleAsync(userID, chatID)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.7 schedule_plan_refine：执行“连续微调排程”graph。
|
||||
if routing.Action == agentrouter.ActionSchedulePlanRefine {
|
||||
reply, refineErr := s.runScheduleRefineFlow(requestCtx, selectedModel, userMessage, userID, chatID, traceID, progress.Emit, outChan, resolvedModelName)
|
||||
if refineErr != nil {
|
||||
// 连续微调失败不再回落普通聊天，直接上报错误。
|
||||
pushErrNonBlocking(errChan, refineErr)
|
||||
return
|
||||
}
|
||||
|
||||
if emitErr := emitSingleAssistantCompletion(outChan, resolvedModelName, reply); emitErr != nil {
|
||||
pushErrNonBlocking(errChan, emitErr)
|
||||
return
|
||||
}
|
||||
requestTotalTokens := snapshotRequestTokenMeter(requestCtx).TotalTokens
|
||||
s.persistChatAfterReply(requestCtx, userID, chatID, userMessage, reply, progress.HistoryText(), progress.DurationSeconds(time.Now()), retryMeta, 0, requestTotalTokens, errChan)
|
||||
s.ensureConversationTitleAsync(userID, chatID)
|
||||
return
|
||||
}
|
||||
|
||||
// 3.8 未知 action 兜底：走普通聊天，保证可用性。
|
||||
s.runNormalChatFlow(requestCtx, selectedModel, resolvedModelName, userMessage, progress.HistoryText(), progress.StartedAt(), retryMeta, ifThinking, userID, chatID, traceID, requestStart, outChan, errChan)
|
||||
}()
|
||||
|
||||
return outChan, errChan
|
||||
}
|
||||
|
||||
@@ -118,7 +118,6 @@ func (s *AgentService) appendConversationHistoryCacheOptimistically(
|
||||
merged = appendConversationHistoryItemIfMissing(merged, item)
|
||||
}
|
||||
sortConversationHistoryItems(merged)
|
||||
merged = attachConversationRetryTotals(merged)
|
||||
|
||||
if err = s.cacheDAO.SetConversationHistoryToCache(ctx, userID, normalizedChatID, merged); err != nil {
|
||||
log.Printf("乐观追加会话历史视图缓存失败 chat_id=%s: %v", normalizedChatID, err)
|
||||
@@ -151,11 +150,9 @@ func buildConversationHistoryItemsFromDB(histories []model.ChatHistory) []model.
|
||||
CreatedAt: history.CreatedAt,
|
||||
ReasoningContent: strings.TrimSpace(derefConversationHistoryText(history.ReasoningContent)),
|
||||
ReasoningDurationSeconds: history.ReasoningDurationSeconds,
|
||||
RetryGroupID: cloneConversationStringPointer(history.RetryGroupID),
|
||||
RetryIndex: cloneConversationIntPointer(history.RetryIndex),
|
||||
})
|
||||
}
|
||||
return attachConversationRetryTotals(items)
|
||||
return items
|
||||
}
|
||||
|
||||
func derefConversationHistoryText(text *string) string {
|
||||
@@ -165,58 +162,6 @@ func derefConversationHistoryText(text *string) string {
|
||||
return *text
|
||||
}
|
||||
|
||||
func attachConversationRetryTotals(items []model.GetConversationHistoryItem) []model.GetConversationHistoryItem {
|
||||
if len(items) == 0 {
|
||||
return items
|
||||
}
|
||||
groupTotals := make(map[string]int)
|
||||
for _, item := range items {
|
||||
if item.RetryGroupID == nil || item.RetryIndex == nil {
|
||||
continue
|
||||
}
|
||||
groupID := strings.TrimSpace(*item.RetryGroupID)
|
||||
if groupID == "" {
|
||||
continue
|
||||
}
|
||||
if *item.RetryIndex > groupTotals[groupID] {
|
||||
groupTotals[groupID] = *item.RetryIndex
|
||||
}
|
||||
}
|
||||
for idx := range items {
|
||||
groupIDPtr := items[idx].RetryGroupID
|
||||
if groupIDPtr == nil {
|
||||
continue
|
||||
}
|
||||
groupID := strings.TrimSpace(*groupIDPtr)
|
||||
total := groupTotals[groupID]
|
||||
if total <= 0 {
|
||||
continue
|
||||
}
|
||||
totalCopy := total
|
||||
items[idx].RetryTotal = &totalCopy
|
||||
}
|
||||
return items
|
||||
}
|
||||
|
||||
func cloneConversationStringPointer(src *string) *string {
|
||||
if src == nil {
|
||||
return nil
|
||||
}
|
||||
text := strings.TrimSpace(*src)
|
||||
if text == "" {
|
||||
return nil
|
||||
}
|
||||
return &text
|
||||
}
|
||||
|
||||
func cloneConversationIntPointer(src *int) *int {
|
||||
if src == nil || *src <= 0 {
|
||||
return nil
|
||||
}
|
||||
value := *src
|
||||
return &value
|
||||
}
|
||||
|
||||
func normalizeConversationHistoryRole(role string) string {
|
||||
switch strings.ToLower(strings.TrimSpace(role)) {
|
||||
case "user":
|
||||
@@ -245,7 +190,6 @@ func buildOptimisticConversationHistoryItem(
|
||||
content string,
|
||||
reasoningContent string,
|
||||
reasoningDurationSeconds int,
|
||||
retryMeta *chatRetryMeta,
|
||||
createdAt time.Time,
|
||||
) model.GetConversationHistoryItem {
|
||||
item := model.GetConversationHistoryItem{
|
||||
@@ -258,11 +202,6 @@ func buildOptimisticConversationHistoryItem(
|
||||
t := createdAt
|
||||
item.CreatedAt = &t
|
||||
}
|
||||
if retryMeta != nil {
|
||||
item.RetryGroupID = retryMeta.GroupIDPtr()
|
||||
item.RetryIndex = retryMeta.IndexPtr()
|
||||
item.RetryTotal = retryMeta.IndexPtr()
|
||||
}
|
||||
return item
|
||||
}
|
||||
|
||||
@@ -284,26 +223,16 @@ func conversationHistoryItemSignature(item model.GetConversationHistoryItem) str
|
||||
return fmt.Sprintf("id:%d", item.ID)
|
||||
}
|
||||
|
||||
groupID := ""
|
||||
if item.RetryGroupID != nil {
|
||||
groupID = strings.TrimSpace(*item.RetryGroupID)
|
||||
}
|
||||
retryIndex := 0
|
||||
if item.RetryIndex != nil {
|
||||
retryIndex = *item.RetryIndex
|
||||
}
|
||||
createdAt := ""
|
||||
if item.CreatedAt != nil {
|
||||
createdAt = item.CreatedAt.UTC().Format(time.RFC3339Nano)
|
||||
}
|
||||
|
||||
return fmt.Sprintf(
|
||||
"%s|%s|%s|%s|%d|%d|%s",
|
||||
"%s|%s|%s|%d|%s",
|
||||
strings.TrimSpace(item.Role),
|
||||
strings.TrimSpace(item.Content),
|
||||
strings.TrimSpace(item.ReasoningContent),
|
||||
groupID,
|
||||
retryIndex,
|
||||
item.ReasoningDurationSeconds,
|
||||
createdAt,
|
||||
)
|
||||
|
||||
@@ -84,12 +84,7 @@ func (s *AgentService) runNewAgentGraph(
|
||||
}
|
||||
}
|
||||
|
||||
// 3. 构建重试元数据。
|
||||
retryMeta, err := s.buildChatRetryMeta(requestCtx, userID, chatID, extra)
|
||||
if err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
return
|
||||
}
|
||||
// 3. retry 机制已下线，不再构建重试元数据。
|
||||
|
||||
// 4. 从 StateStore 加载或创建 RuntimeState。
|
||||
// 恢复场景（confirm/ask_user）同时拿到快照中保存的 ConversationContext，
|
||||
@@ -137,6 +132,31 @@ func (s *AgentService) runNewAgentGraph(
|
||||
}
|
||||
}
|
||||
|
||||
cs = runtimeState.EnsureCommonState()
|
||||
|
||||
// 5.7 先把本轮用户输入落库，确保后续可见 assistant 消息按真实时间线追加。
|
||||
userMsg := schema.UserMessage(userMessage)
|
||||
if err := s.persistNewAgentConversationMessage(requestCtx, userID, chatID, userMsg, 0); err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
return
|
||||
}
|
||||
|
||||
persistVisibleMessage := func(persistCtx context.Context, state *newagentmodel.CommonState, msg *schema.Message) error {
|
||||
targetState := state
|
||||
if targetState == nil {
|
||||
targetState = runtimeState.EnsureCommonState()
|
||||
}
|
||||
if targetState != nil {
|
||||
if targetState.UserID <= 0 {
|
||||
targetState.UserID = userID
|
||||
}
|
||||
if strings.TrimSpace(targetState.ConversationID) == "" {
|
||||
targetState.ConversationID = chatID
|
||||
}
|
||||
}
|
||||
return s.persistNewAgentConversationMessage(persistCtx, userID, chatID, msg, 0)
|
||||
}
|
||||
|
||||
// 6. 构造 AgentGraphRequest。
|
||||
var confirmAction string
|
||||
if len(extra) > 0 {
|
||||
@@ -163,22 +183,23 @@ func (s *AgentService) runNewAgentGraph(
|
||||
|
||||
// 9. 构造 AgentGraphDeps（由 cmd/start.go 注入的依赖）。
|
||||
deps := newagentmodel.AgentGraphDeps{
|
||||
ChatClient: chatClient,
|
||||
PlanClient: planClient,
|
||||
ExecuteClient: executeClient,
|
||||
DeliverClient: deliverClient,
|
||||
ChunkEmitter: chunkEmitter,
|
||||
StateStore: s.agentStateStore,
|
||||
ToolRegistry: s.toolRegistry,
|
||||
ScheduleProvider: s.scheduleProvider,
|
||||
SchedulePersistor: s.schedulePersistor,
|
||||
CompactionStore: s.compactionStore,
|
||||
RoughBuildFunc: s.makeRoughBuildFunc(),
|
||||
WriteSchedulePreview: s.makeWriteSchedulePreviewFunc(),
|
||||
MemoryFuture: memoryFuture,
|
||||
ThinkingPlan: viper.GetBool("agent.thinking.plan"),
|
||||
ThinkingExecute: viper.GetBool("agent.thinking.execute"),
|
||||
ThinkingDeliver: viper.GetBool("agent.thinking.deliver"),
|
||||
ChatClient: chatClient,
|
||||
PlanClient: planClient,
|
||||
ExecuteClient: executeClient,
|
||||
DeliverClient: deliverClient,
|
||||
ChunkEmitter: chunkEmitter,
|
||||
StateStore: s.agentStateStore,
|
||||
ToolRegistry: s.toolRegistry,
|
||||
ScheduleProvider: s.scheduleProvider,
|
||||
SchedulePersistor: s.schedulePersistor,
|
||||
CompactionStore: s.compactionStore,
|
||||
RoughBuildFunc: s.makeRoughBuildFunc(),
|
||||
WriteSchedulePreview: s.makeWriteSchedulePreviewFunc(),
|
||||
MemoryFuture: memoryFuture,
|
||||
ThinkingPlan: viper.GetBool("agent.thinking.plan"),
|
||||
ThinkingExecute: viper.GetBool("agent.thinking.execute"),
|
||||
ThinkingDeliver: viper.GetBool("agent.thinking.deliver"),
|
||||
PersistVisibleMessage: persistVisibleMessage,
|
||||
}
|
||||
|
||||
// 10. 构造 AgentGraphRunInput 并运行 graph。
|
||||
@@ -197,12 +218,13 @@ func (s *AgentService) runNewAgentGraph(
|
||||
pushErrNonBlocking(errChan, fmt.Errorf("graph 执行失败: %w", graphErr))
|
||||
|
||||
// Graph 出错时回退普通聊天，保证可用性。回退使用 Pro 模型。
|
||||
s.runNormalChatFlow(requestCtx, s.AIHub.Pro, resolvedModelName, userMessage, "", nil, retryMeta, thinkingModeToBool(thinkingMode), userID, chatID, traceID, requestStart, outChan, errChan)
|
||||
s.runNormalChatFlow(requestCtx, s.AIHub.Pro, resolvedModelName, userMessage, true, "", nil, thinkingModeToBool(thinkingMode), userID, chatID, traceID, requestStart, outChan, errChan)
|
||||
return
|
||||
}
|
||||
|
||||
// 11. 持久化聊天历史（用户消息 + 助手回复）。
|
||||
s.persistChatAfterGraph(requestCtx, userID, chatID, userMessage, finalState, retryMeta, requestStart, outChan, errChan)
|
||||
requestTotalTokens := snapshotRequestTokenMeter(requestCtx).TotalTokens
|
||||
s.adjustNewAgentRequestTokenUsage(requestCtx, userID, chatID, requestTotalTokens)
|
||||
// 11.5. 将最终状态快照异步写入 MySQL（通过 outbox）。
|
||||
// Deliver 节点已将快照保存到 Redis（2h TTL），此处通过 outbox 异步写入 MySQL 做永久存储。
|
||||
if finalState != nil {
|
||||
@@ -369,135 +391,89 @@ func (s *AgentService) loadConversationContext(ctx context.Context, chatID, user
|
||||
return conversationContext
|
||||
}
|
||||
|
||||
// persistChatAfterGraph graph 执行完成后持久化聊天历史。
|
||||
func (s *AgentService) persistChatAfterGraph(
|
||||
// persistNewAgentConversationMessage 负责把 newAgent 链路里"真正对用户可见"的消息统一落到 Redis + MySQL。
|
||||
//
|
||||
// 职责边界：
|
||||
// 1. 只做单条消息的持久化，不做 graph 流程控制；
|
||||
// 2. TokensConsumed 由调用方显式传入，newAgent 逐条可见消息默认写 0；
|
||||
// 3. Redis 失败只记日志，DB 失败返回错误，便于调用方决定是否中止当前链路。
|
||||
func (s *AgentService) persistNewAgentConversationMessage(
|
||||
ctx context.Context,
|
||||
userID int,
|
||||
chatID string,
|
||||
userMessage string,
|
||||
finalState *newagentmodel.AgentGraphState,
|
||||
retryMeta *chatRetryMeta,
|
||||
requestStart time.Time,
|
||||
outChan chan<- string,
|
||||
errChan chan error,
|
||||
) {
|
||||
if finalState == nil {
|
||||
return
|
||||
msg *schema.Message,
|
||||
tokensConsumed int,
|
||||
) error {
|
||||
if s == nil || msg == nil {
|
||||
return nil
|
||||
}
|
||||
role := strings.TrimSpace(string(msg.Role))
|
||||
content := strings.TrimSpace(msg.Content)
|
||||
if role == "" || content == "" {
|
||||
return nil
|
||||
}
|
||||
if userID <= 0 || strings.TrimSpace(chatID) == "" {
|
||||
return fmt.Errorf("newAgent visible message persist: invalid conversation identity")
|
||||
}
|
||||
if ctx == nil {
|
||||
ctx = context.Background()
|
||||
}
|
||||
|
||||
// 1. 持久化用户消息：先写 LLM 上下文 Redis，再落 DB，最后更新 UI 历史缓存。
|
||||
userMsg := &schema.Message{Role: schema.User, Content: userMessage}
|
||||
if retryExtra := retryMeta.CacheExtra(); len(retryExtra) > 0 {
|
||||
userMsg.Extra = retryExtra
|
||||
persistMsg := &schema.Message{
|
||||
Role: msg.Role,
|
||||
Content: content,
|
||||
ReasoningContent: strings.TrimSpace(msg.ReasoningContent),
|
||||
}
|
||||
if err := s.agentCache.PushMessage(ctx, chatID, userMsg); err != nil {
|
||||
log.Printf("写入用户消息到 LLM 上下文 Redis 失败 chat=%s: %v", chatID, err)
|
||||
if len(msg.Extra) > 0 {
|
||||
persistMsg.Extra = make(map[string]any, len(msg.Extra))
|
||||
for key, value := range msg.Extra {
|
||||
persistMsg.Extra[key] = value
|
||||
}
|
||||
}
|
||||
|
||||
userPayload := model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "user",
|
||||
Message: userMessage,
|
||||
ReasoningContent: "",
|
||||
ReasoningDurationSeconds: 0,
|
||||
RetryGroupID: retryMeta.GroupIDPtr(),
|
||||
RetryIndex: retryMeta.IndexPtr(),
|
||||
RetryFromUserMessageID: retryMeta.FromUserMessageIDPtr(),
|
||||
RetryFromAssistantMessageID: retryMeta.FromAssistantMessageIDPtr(),
|
||||
TokensConsumed: 0,
|
||||
if err := s.agentCache.PushMessage(ctx, chatID, persistMsg); err != nil {
|
||||
log.Printf("写入 newAgent 可见消息到 Redis 失败 chat=%s role=%s: %v", chatID, role, err)
|
||||
}
|
||||
if err := s.PersistChatHistory(ctx, userPayload); err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
|
||||
reasoningDurationSeconds := 0
|
||||
if persistMsg.Extra != nil {
|
||||
switch v := persistMsg.Extra["reasoning_duration_seconds"].(type) {
|
||||
case int:
|
||||
reasoningDurationSeconds = v
|
||||
case int64:
|
||||
reasoningDurationSeconds = int(v)
|
||||
case float64:
|
||||
reasoningDurationSeconds = int(v)
|
||||
}
|
||||
}
|
||||
userCreatedAt := time.Now()
|
||||
|
||||
persistPayload := model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: role,
|
||||
Message: content,
|
||||
ReasoningContent: strings.TrimSpace(persistMsg.ReasoningContent),
|
||||
ReasoningDurationSeconds: reasoningDurationSeconds,
|
||||
TokensConsumed: tokensConsumed,
|
||||
}
|
||||
if err := s.PersistChatHistory(ctx, persistPayload); err != nil {
|
||||
return err
|
||||
}
|
||||
|
||||
now := time.Now()
|
||||
s.appendConversationHistoryCacheOptimistically(
|
||||
context.Background(),
|
||||
ctx,
|
||||
userID,
|
||||
chatID,
|
||||
buildOptimisticConversationHistoryItem("user", userMessage, "", 0, retryMeta, userCreatedAt),
|
||||
buildOptimisticConversationHistoryItem(
|
||||
role,
|
||||
content,
|
||||
persistPayload.ReasoningContent,
|
||||
reasoningDurationSeconds,
|
||||
now,
|
||||
),
|
||||
)
|
||||
|
||||
// 2. 从 ConversationContext 提取助手回复（最后一条 assistant 消息）。
|
||||
conversationContext := finalState.ConversationContext
|
||||
if conversationContext == nil || len(conversationContext.History) == 0 {
|
||||
return
|
||||
}
|
||||
|
||||
var lastAssistantMsg *schema.Message
|
||||
for i := len(conversationContext.History) - 1; i >= 0; i-- {
|
||||
msg := conversationContext.History[i]
|
||||
if msg.Role == schema.Assistant {
|
||||
lastAssistantMsg = msg
|
||||
break
|
||||
}
|
||||
}
|
||||
|
||||
if lastAssistantMsg == nil {
|
||||
return
|
||||
}
|
||||
|
||||
assistantReply := lastAssistantMsg.Content
|
||||
reasoningContent := lastAssistantMsg.ReasoningContent
|
||||
var reasoningDurationSeconds int
|
||||
if lastAssistantMsg.Extra != nil {
|
||||
if dur, ok := lastAssistantMsg.Extra["reasoning_duration_seconds"].(float64); ok {
|
||||
reasoningDurationSeconds = int(dur)
|
||||
}
|
||||
}
|
||||
|
||||
// 3. 持久化助手消息：先写 LLM 上下文 Redis，再落 DB，最后更新 UI 历史缓存。
|
||||
assistantMsg := &schema.Message{
|
||||
Role: schema.Assistant,
|
||||
Content: assistantReply,
|
||||
ReasoningContent: reasoningContent,
|
||||
}
|
||||
if reasoningDurationSeconds > 0 {
|
||||
assistantMsg.Extra = map[string]any{"reasoning_duration_seconds": reasoningDurationSeconds}
|
||||
}
|
||||
if retryExtra := retryMeta.CacheExtra(); len(retryExtra) > 0 {
|
||||
if assistantMsg.Extra == nil {
|
||||
assistantMsg.Extra = make(map[string]any)
|
||||
}
|
||||
for k, v := range retryExtra {
|
||||
assistantMsg.Extra[k] = v
|
||||
}
|
||||
}
|
||||
if err := s.agentCache.PushMessage(context.Background(), chatID, assistantMsg); err != nil {
|
||||
log.Printf("写入助手消息到 LLM 上下文 Redis 失败 chat=%s: %v", chatID, err)
|
||||
}
|
||||
|
||||
requestTotalTokens := snapshotRequestTokenMeter(ctx).TotalTokens
|
||||
assistantPayload := model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "assistant",
|
||||
Message: assistantReply,
|
||||
ReasoningContent: reasoningContent,
|
||||
ReasoningDurationSeconds: reasoningDurationSeconds,
|
||||
RetryGroupID: retryMeta.GroupIDPtr(),
|
||||
RetryIndex: retryMeta.IndexPtr(),
|
||||
RetryFromUserMessageID: retryMeta.FromUserMessageIDPtr(),
|
||||
RetryFromAssistantMessageID: retryMeta.FromAssistantMessageIDPtr(),
|
||||
TokensConsumed: requestTotalTokens,
|
||||
}
|
||||
if err := s.PersistChatHistory(ctx, assistantPayload); err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
} else {
|
||||
s.appendConversationHistoryCacheOptimistically(
|
||||
context.Background(),
|
||||
userID,
|
||||
chatID,
|
||||
buildOptimisticConversationHistoryItem(
|
||||
"assistant",
|
||||
assistantReply,
|
||||
reasoningContent,
|
||||
reasoningDurationSeconds,
|
||||
retryMeta,
|
||||
time.Now(),
|
||||
),
|
||||
)
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
// makeRoughBuildFunc 把 AgentService 上的 HybridScheduleWithPlanMultiFunc 封装成
|
||||
@@ -509,6 +485,38 @@ func (s *AgentService) persistChatAfterGraph(
|
||||
// placement，普通时段放置的任务全部被丢弃。
|
||||
// 正确做法：使用第一个返回值 []HybridScheduleEntry，过滤 Status="suggested" 且 TaskItemID>0 的条目，
|
||||
// 这样嵌入和非嵌入的粗排结果都能正确写入 ScheduleState。
|
||||
// adjustNewAgentRequestTokenUsage 负责把本轮 graph 的请求级 token 一次性回写到账本。
|
||||
//
|
||||
// 说明：
|
||||
// 1. newAgent 逐条可见消息都按 0 token 落库，最终统一在这里补记整轮消耗；
|
||||
// 2. 如果启用了 outbox，就沿用异步 token 调整事件，保持写账口径一致；
|
||||
// 3. 该步骤属于请求收尾，不应反过来打断用户已看到的回复。
|
||||
func (s *AgentService) adjustNewAgentRequestTokenUsage(ctx context.Context, userID int, chatID string, deltaTokens int) {
|
||||
if s == nil || userID <= 0 || strings.TrimSpace(chatID) == "" || deltaTokens <= 0 {
|
||||
return
|
||||
}
|
||||
if ctx == nil {
|
||||
ctx = context.Background()
|
||||
}
|
||||
|
||||
if s.eventPublisher != nil {
|
||||
if err := eventsvc.PublishChatTokenUsageAdjustRequested(ctx, s.eventPublisher, model.ChatTokenUsageAdjustPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
TokensDelta: deltaTokens,
|
||||
Reason: "new_agent_request",
|
||||
TriggeredAt: time.Now(),
|
||||
}); err != nil {
|
||||
log.Printf("写入 newAgent 请求级 token 调整事件失败 chat=%s tokens=%d err=%v", chatID, deltaTokens, err)
|
||||
}
|
||||
return
|
||||
}
|
||||
|
||||
if err := s.repo.AdjustTokenUsage(ctx, userID, chatID, deltaTokens); err != nil {
|
||||
log.Printf("同步写入 newAgent 请求级 token 调整失败 chat=%s tokens=%d err=%v", chatID, deltaTokens, err)
|
||||
}
|
||||
}
|
||||
|
||||
func (s *AgentService) makeRoughBuildFunc() newagentmodel.RoughBuildFunc {
|
||||
if s.HybridScheduleWithPlanMultiFunc == nil {
|
||||
return nil
|
||||
|
||||
@@ -15,7 +15,6 @@ import (
|
||||
agentstream "github.com/LoveLosita/smartflow/backend/agent/stream"
|
||||
"github.com/LoveLosita/smartflow/backend/model"
|
||||
"github.com/cloudwego/eino-ext/components/model/ark"
|
||||
"github.com/cloudwego/eino/schema"
|
||||
"github.com/google/uuid"
|
||||
)
|
||||
|
||||
@@ -302,109 +301,3 @@ func (s *AgentService) decideQuickNoteRouting(ctx context.Context, selectedModel
|
||||
_ = s
|
||||
return agentrouter.DecideQuickNoteRouting(ctx, selectedModel, userMessage)
|
||||
}
|
||||
|
||||
// persistChatAfterReply 在“随口记 graph”返回后，复用当前项目的后置持久化策略：
|
||||
// 1) 用户消息写 Redis + outbox/DB；
|
||||
// 2) 助手消息写 Redis + outbox/DB。
|
||||
func (s *AgentService) persistChatAfterReply(
|
||||
ctx context.Context,
|
||||
userID int,
|
||||
chatID string,
|
||||
userMessage string,
|
||||
assistantReply string,
|
||||
assistantReasoning string,
|
||||
assistantReasoningDurationSeconds int,
|
||||
retryMeta *chatRetryMeta,
|
||||
userTokens int,
|
||||
assistantTokens int,
|
||||
errChan chan error,
|
||||
) {
|
||||
// 1. 先把用户消息写入 Redis，保证会话上下文“马上可见”。
|
||||
userMsg := &schema.Message{Role: schema.User, Content: userMessage}
|
||||
if retryExtra := retryMeta.CacheExtra(); len(retryExtra) > 0 {
|
||||
userMsg.Extra = retryExtra
|
||||
}
|
||||
if err := s.agentCache.PushMessage(ctx, chatID, userMsg); err != nil {
|
||||
log.Printf("写入用户消息到 Redis 失败: %v", err)
|
||||
}
|
||||
|
||||
// 2. 再把用户消息写入可靠持久化通道（outbox 或同步 DB）。
|
||||
if err := s.PersistChatHistory(ctx, model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "user",
|
||||
Message: userMessage,
|
||||
ReasoningContent: "",
|
||||
ReasoningDurationSeconds: 0,
|
||||
RetryGroupID: retryMeta.GroupIDPtr(),
|
||||
RetryIndex: retryMeta.IndexPtr(),
|
||||
RetryFromUserMessageID: retryMeta.FromUserMessageIDPtr(),
|
||||
RetryFromAssistantMessageID: retryMeta.FromAssistantMessageIDPtr(),
|
||||
TokensConsumed: userTokens,
|
||||
}); err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
return
|
||||
}
|
||||
userCreatedAt := time.Now()
|
||||
s.appendConversationHistoryCacheOptimistically(
|
||||
context.Background(),
|
||||
userID,
|
||||
chatID,
|
||||
buildOptimisticConversationHistoryItem(
|
||||
"user",
|
||||
userMessage,
|
||||
"",
|
||||
0,
|
||||
retryMeta,
|
||||
userCreatedAt,
|
||||
),
|
||||
)
|
||||
|
||||
// 3. 助手消息同样遵循“Redis 先行 + 可靠持久化补齐”策略。
|
||||
assistantMsg := &schema.Message{Role: schema.Assistant, Content: assistantReply, ReasoningContent: assistantReasoning}
|
||||
if assistantReasoningDurationSeconds > 0 {
|
||||
assistantMsg.Extra = map[string]any{"reasoning_duration_seconds": assistantReasoningDurationSeconds}
|
||||
}
|
||||
if retryExtra := retryMeta.CacheExtra(); len(retryExtra) > 0 {
|
||||
if assistantMsg.Extra == nil {
|
||||
assistantMsg.Extra = make(map[string]any, len(retryExtra))
|
||||
}
|
||||
for key, value := range retryExtra {
|
||||
assistantMsg.Extra[key] = value
|
||||
}
|
||||
}
|
||||
if err := s.agentCache.PushMessage(context.Background(), chatID, assistantMsg); err != nil {
|
||||
log.Printf("写入助手消息到 Redis 失败: %v", err)
|
||||
}
|
||||
|
||||
// 4. 助手消息持久化失败不阻断主流程，通过 errChan 异步上报。
|
||||
if err := s.PersistChatHistory(context.Background(), model.ChatHistoryPersistPayload{
|
||||
UserID: userID,
|
||||
ConversationID: chatID,
|
||||
Role: "assistant",
|
||||
Message: assistantReply,
|
||||
ReasoningContent: assistantReasoning,
|
||||
ReasoningDurationSeconds: assistantReasoningDurationSeconds,
|
||||
RetryGroupID: retryMeta.GroupIDPtr(),
|
||||
RetryIndex: retryMeta.IndexPtr(),
|
||||
RetryFromUserMessageID: retryMeta.FromUserMessageIDPtr(),
|
||||
RetryFromAssistantMessageID: retryMeta.FromAssistantMessageIDPtr(),
|
||||
TokensConsumed: assistantTokens,
|
||||
}); err != nil {
|
||||
pushErrNonBlocking(errChan, err)
|
||||
return
|
||||
}
|
||||
s.appendConversationHistoryCacheOptimistically(
|
||||
context.Background(),
|
||||
userID,
|
||||
chatID,
|
||||
buildOptimisticConversationHistoryItem(
|
||||
"assistant",
|
||||
assistantReply,
|
||||
assistantReasoning,
|
||||
assistantReasoningDurationSeconds,
|
||||
retryMeta,
|
||||
userCreatedAt.Add(time.Millisecond),
|
||||
),
|
||||
)
|
||||
}
|
||||
|
||||
@@ -71,10 +71,6 @@ func RegisterChatHistoryPersistHandler(
|
||||
payload.Message,
|
||||
payload.ReasoningContent,
|
||||
payload.ReasoningDurationSeconds,
|
||||
payload.RetryGroupID,
|
||||
payload.RetryIndex,
|
||||
payload.RetryFromUserMessageID,
|
||||
payload.RetryFromAssistantMessageID,
|
||||
payload.TokensConsumed,
|
||||
); err != nil {
|
||||
return err
|
||||
|
||||
Reference in New Issue
Block a user