feat: add context compaction to automatically summarize old messages (#compaction) (#276)

When input tokens exceed a configurable threshold after a conversation round,
the system asynchronously compacts older messages into a summary. Cascading
compactions reference prior summaries via <prior_context> tags to maintain
conversational continuity without duplicating content.

- Add bot_history_message_compacts table and compact_id on messages
- Add compaction_enabled, compaction_threshold, compaction_model_id to bots
- Implement compaction service (internal/compaction) with LLM summarization
- Integrate into conversation flow: replace compacted messages with summaries
  wrapped in <summary> tags during context loading
- Add REST API endpoints (GET/DELETE /bots/:bot_id/compaction/logs)
- Add frontend Compaction tab with settings and log viewer
- Wire compaction service into both dev (cmd/agent) and prod (cmd/memoh) entry points
- Update test mocks to include new GetBotByID columns
This commit is contained in:
Acbox Liu
2026-03-22 14:26:00 +08:00
committed by GitHub
parent 91e5e44509
commit de62f94315
40 changed files with 2375 additions and 197 deletions
+5 -1
View File
@@ -39,6 +39,7 @@ import (
"github.com/memohai/memoh/internal/channel/inbound"
"github.com/memohai/memoh/internal/channel/route"
"github.com/memohai/memoh/internal/command"
"github.com/memohai/memoh/internal/compaction"
"github.com/memohai/memoh/internal/config"
ctr "github.com/memohai/memoh/internal/containerd"
"github.com/memohai/memoh/internal/conversation"
@@ -209,6 +210,7 @@ func runServe() {
schedule.NewService,
provideHeartbeatTriggerer,
heartbeat.NewService,
compaction.NewService,
// containerd handler & tool gateway
provideContainerdHandler,
@@ -231,6 +233,7 @@ func runServe() {
provideServerHandler(handlers.NewBindHandler),
provideServerHandler(handlers.NewScheduleHandler),
provideServerHandler(handlers.NewHeartbeatHandler),
provideServerHandler(handlers.NewCompactionHandler),
provideServerHandler(handlers.NewSubagentHandler),
provideServerHandler(handlers.NewChannelHandler),
provideServerHandler(feishu.NewWebhookServerHandler),
@@ -432,13 +435,14 @@ func injectToolProviders(a *agentpkg.Agent, providers []agenttools.ToolProvider)
a.SetToolProviders(providers)
}
func provideChatResolver(log *slog.Logger, a *agentpkg.Agent, modelsService *models.Service, queries *dbsqlc.Queries, chatService *conversation.Service, msgService *message.DBService, settingsService *settings.Service, mediaService *media.Service, containerdHandler *handlers.ContainerdHandler, memoryRegistry *memprovider.Registry, sessionService *sessionpkg.Service, eventHub *event.Hub) *flow.Resolver {
func provideChatResolver(log *slog.Logger, a *agentpkg.Agent, modelsService *models.Service, queries *dbsqlc.Queries, chatService *conversation.Service, msgService *message.DBService, settingsService *settings.Service, mediaService *media.Service, containerdHandler *handlers.ContainerdHandler, memoryRegistry *memprovider.Registry, sessionService *sessionpkg.Service, eventHub *event.Hub, compactionService *compaction.Service) *flow.Resolver {
resolver := flow.NewResolver(log, modelsService, queries, chatService, msgService, settingsService, a, 120*time.Second)
resolver.SetMemoryRegistry(memoryRegistry)
resolver.SetSkillLoader(&skillLoaderAdapter{handler: containerdHandler})
resolver.SetGatewayAssetLoader(&gatewayAssetLoaderAdapter{media: mediaService})
resolver.SetSessionService(sessionService)
resolver.SetEventPublisher(eventHub)
resolver.SetCompactionService(compactionService)
return resolver
}
+5 -1
View File
@@ -40,6 +40,7 @@ import (
"github.com/memohai/memoh/internal/channel/inbound"
"github.com/memohai/memoh/internal/channel/route"
"github.com/memohai/memoh/internal/command"
"github.com/memohai/memoh/internal/compaction"
"github.com/memohai/memoh/internal/config"
ctr "github.com/memohai/memoh/internal/containerd"
"github.com/memohai/memoh/internal/conversation"
@@ -139,6 +140,7 @@ func runServe() {
schedule.NewService,
provideHeartbeatTriggerer,
heartbeat.NewService,
compaction.NewService,
provideContainerdHandler,
provideFederationGateway,
provideToolGatewayService,
@@ -157,6 +159,7 @@ func runServe() {
provideServerHandler(handlers.NewBindHandler),
provideServerHandler(handlers.NewScheduleHandler),
provideServerHandler(handlers.NewHeartbeatHandler),
provideServerHandler(handlers.NewCompactionHandler),
provideServerHandler(handlers.NewSubagentHandler),
provideServerHandler(handlers.NewChannelHandler),
provideServerHandler(feishu.NewWebhookServerHandler),
@@ -340,13 +343,14 @@ func injectToolProviders(a *agentpkg.Agent, providers []agenttools.ToolProvider)
a.SetToolProviders(providers)
}
func provideChatResolver(log *slog.Logger, a *agentpkg.Agent, modelsService *models.Service, queries *dbsqlc.Queries, chatService *conversation.Service, msgService *message.DBService, settingsService *settings.Service, mediaService *media.Service, containerdHandler *handlers.ContainerdHandler, memoryRegistry *memprovider.Registry, sessionService *sessionpkg.Service, eventHub *event.Hub) *flow.Resolver {
func provideChatResolver(log *slog.Logger, a *agentpkg.Agent, modelsService *models.Service, queries *dbsqlc.Queries, chatService *conversation.Service, msgService *message.DBService, settingsService *settings.Service, mediaService *media.Service, containerdHandler *handlers.ContainerdHandler, memoryRegistry *memprovider.Registry, sessionService *sessionpkg.Service, eventHub *event.Hub, compactionService *compaction.Service) *flow.Resolver {
resolver := flow.NewResolver(log, modelsService, queries, chatService, msgService, settingsService, a, 120*time.Second)
resolver.SetMemoryRegistry(memoryRegistry)
resolver.SetSkillLoader(&skillLoaderAdapter{handler: containerdHandler})
resolver.SetGatewayAssetLoader(&gatewayAssetLoaderAdapter{media: mediaService})
resolver.SetSessionService(sessionService)
resolver.SetEventPublisher(eventHub)
resolver.SetCompactionService(compactionService)
return resolver
}