fix(agent): worker 跨 work 回看工具被旧迁移误删，改绑定并补绑回

seedWorkerReadToolsUnbind 的解绑清单里仍列着 search_all_worker_traces /
get_worker_trace，而 d531ea9 已把这两个加进 WorkerTools()——新库启动时先
seed 进去又被这段迁移撸掉，导致 worker（含 benchmark 镜像）实际拿不到。

- 从解绑清单去掉这两个，只保留 list_facts/node_detail/list_companies/list_worker_traces
- 新增一次性 seedWorkerCrossTraceRebind：把这两个补绑回 worker，修复已跑过旧
  v1 迁移、被误删的现存库；新库为无害 no-op

server/orchestration.go

@@ -490,7 +490,8 @@ func (s *Server) seedOrchestrationTools() {
 	s.seedPlannerDefaultBindings()
 	s.seedPlannerListAssetsBinding()
 	s.seedCompanyScopeRebind()
-	s.seedWorkerReadToolsUnbind() // list_facts/node_detail/list_companies/跨 work 检索从 worker 默认解绑(一次性)
+	s.seedWorkerReadToolsUnbind()  // list_facts/node_detail/list_companies/list_worker_traces 从 worker 默认解绑(一次性)
+	s.seedWorkerCrossTraceRebind() // 修复旧 v1 迁移误删：把 search_all_worker_traces/get_worker_trace 补绑回 worker(一次性)
 	s.seedAutoReportFindingBinding()
 	s.unbindGoalMetDefault()
 	s.reseedGoalsPrompt()             // goals 提示词加入「抽操作约束」步 → 旧库追加一版新默认(一次性)
@@ -846,21 +847,25 @@ func (s *Server) seedCompanyScopeRebind() {
 	_ = s.m.pg.SetSetting(flag, "true")
 }
 
-// seedWorkerReadToolsUnbind strips the read-context / cross-work tools off worker's
-// default binding ONCE on existing DBs (guarded by a settings flag): a worker executes
-// one intent and writes back — reading facts/nodes/companies and pulling other workers'
-// traces is a planning/main concern, not the executor's. Fresh DBs already lack these via
-// WorkerTools(); this only backfills old rows without overriding a user who deliberately
-// re-binds worker. Each RemoveAgentFromTool is per-tool + membership-guarded, so
-// planner/mainagent bindings of the same tool are untouched.
+// seedWorkerReadToolsUnbind strips the read-context tools off worker's default
+// binding ONCE on existing DBs (guarded by a settings flag): a worker executes one
+// intent and writes back — reading facts/nodes/companies and listing all workers'
+// traces is a planning/main concern, not the executor's. Fresh DBs already lack these
+// via WorkerTools(); this only backfills old rows without overriding a user who
+// deliberately re-binds worker. Each RemoveAgentFromTool is per-tool +
+// membership-guarded, so planner/mainagent bindings of the same tool are untouched.
+//
+// NOTE: search_all_worker_traces / get_worker_trace are intentionally NOT unbound —
+// worker owns them for cross-work look-back (see WorkerTools). They used to be in this
+// list back when worker lacked them; seedWorkerCrossTraceRebind repairs DBs whose old
+// v1 run stripped them.
 func (s *Server) seedWorkerReadToolsUnbind() {
 	const flag = "worker_readtools_unbind_v1"
 	if v, _, _ := s.m.pg.GetSetting(flag); v == "true" {
 		return
 	}
 	for _, k := range []string{
-		"list_facts", "node_detail", "list_companies",
-		"search_all_worker_traces", "list_worker_traces", "get_worker_trace",
+		"list_facts", "node_detail", "list_companies", "list_worker_traces",
 	} {
 		if err := s.m.pg.RemoveAgentFromTool("worker", k); err != nil {
 			log.Printf("[worker] %s 从 worker 解绑失败: %v", k, err)
@@ -870,6 +875,26 @@ func (s *Server) seedWorkerReadToolsUnbind() {
 	_ = s.m.pg.SetSetting(flag, "true")
 }
 
+// seedWorkerCrossTraceRebind re-binds the cross-work look-back tools onto worker ONCE
+// (guarded by a settings flag): an earlier seedWorkerReadToolsUnbind (v1) wrongly
+// stripped search_all_worker_traces / get_worker_trace from worker after they had been
+// added to WorkerTools(), so any DB that ran that migration lost them. Fresh DBs already
+// have them via WorkerTools() and this is a harmless no-op there. One-shot + flag-guarded
+// so a user who later deliberately unbinds them isn't overridden.
+func (s *Server) seedWorkerCrossTraceRebind() {
+	const flag = "worker_crosstrace_rebind_v1"
+	if v, _, _ := s.m.pg.GetSetting(flag); v == "true" {
+		return
+	}
+	if err := s.m.pg.AddAgentToToolBinding("worker", []string{
+		"search_all_worker_traces", "get_worker_trace",
+	}); err != nil {
+		log.Printf("[worker] 跨 work 回看工具补绑失败: %v", err)
+		return // 出错则不落 flag，下次启动重试
+	}
+	_ = s.m.pg.SetSetting(flag, "true")
+}
+
 // seedAutoDefaultBindings adds "auto" to the task-op + platform tools' bindings
 // ONCE (guarded by a settings flag), so existing DBs whose tool rows were seeded
 // before Auto existed still give Auto its default toolset — without re-adding it