feat: configure acquire chat batch size (#23196)

## Summary
- add a hidden deployment config option for chat acquire batch size
(`CODER_CHAT_ACQUIRE_BATCH_SIZE` / `chat.acquireBatchSize`)
- thread the configured value into chatd startup while preserving the
existing default of `10`
- clamp the deployment value to the `int32` range before passing it into
chatd
- regenerate the API/docs/types/testdata artifacts for the new config
field

## Why
`chatd` currently acquires pending chats in batches of `10` via a
compile-time default. This change makes that batch size
operator-configurable from deployment config, so we can tune acquisition
behavior without another code change.

coderd/apidoc/docs.go

@@ -12714,6 +12714,9 @@ const docTemplate = `{
                 },
                 "bridge": {
                     "$ref": "#/definitions/codersdk.AIBridgeConfig"
+                },
+                "chat": {
+                    "$ref": "#/definitions/codersdk.ChatConfig"
                 }
             }
         },
@@ -13771,6 +13774,14 @@ const docTemplate = `{
                 }
             }
         },
+        "codersdk.ChatConfig": {
+            "type": "object",
+            "properties": {
+                "acquire_batch_size": {
+                    "type": "integer"
+                }
+            }
+        },
         "codersdk.ConnectionLatency": {
             "type": "object",
             "properties": {

coderd/apidoc/swagger.json

@@ -11320,6 +11320,9 @@
 				},
 				"bridge": {
 					"$ref": "#/definitions/codersdk.AIBridgeConfig"
+				},
+				"chat": {
+					"$ref": "#/definitions/codersdk.ChatConfig"
 				}
 			}
 		},
@@ -12342,6 +12345,14 @@
 				}
 			}
 		},
+		"codersdk.ChatConfig": {
+			"type": "object",
+			"properties": {
+				"acquire_batch_size": {
+					"type": "integer"
+				}
+			}
+		},
 		"codersdk.ConnectionLatency": {
 			"type": "object",
 			"properties": {

coderd/chatd/chatd.go

@@ -58,11 +58,11 @@ const (
 	// of 5 means recovery runs at 1/5 of the stale-after duration.
 	staleRecoveryIntervalDivisor = 5
 
-	// maxChatsPerAcquire is the maximum number of chats to
+	// DefaultMaxChatsPerAcquire is the maximum number of chats to
 	// acquire in a single processOnce call. Batching avoids
 	// waiting a full polling interval between acquisitions
 	// when many chats are pending.
-	maxChatsPerAcquire int32 = 10
+	DefaultMaxChatsPerAcquire int32 = 10
 
 	defaultSubagentInstruction = "You are running as a delegated sub-agent chat. Complete the delegated task and provide clear, concise assistant responses for the parent agent."
 )
@@ -98,6 +98,7 @@ type Server struct {
 
 	// Configuration
 	pendingChatAcquireInterval time.Duration
+	maxChatsPerAcquire         int32
 	inFlightChatStaleAfter     time.Duration
 }
 
@@ -1174,6 +1175,7 @@ type Config struct {
 	ReplicaID                  uuid.UUID
 	SubscribeFn                SubscribeFn
 	PendingChatAcquireInterval time.Duration
+	MaxChatsPerAcquire         int32
 	InFlightChatStaleAfter     time.Duration
 	AgentConn                  AgentConnFunc
 	CreateWorkspace            chattool.CreateWorkspaceFn
@@ -1199,6 +1201,11 @@ func New(cfg Config) *Server {
 		inFlightChatStaleAfter = DefaultInFlightChatStaleAfter
 	}
 
+	maxChatsPerAcquire := cfg.MaxChatsPerAcquire
+	if maxChatsPerAcquire <= 0 {
+		maxChatsPerAcquire = DefaultMaxChatsPerAcquire
+	}
+
 	workerID := cfg.ReplicaID
 	if workerID == uuid.Nil {
 		workerID = uuid.New()
@@ -1219,6 +1226,7 @@ func New(cfg Config) *Server {
 		providerAPIKeys:            cfg.ProviderAPIKeys,
 		instructionCache:           make(map[uuid.UUID]cachedInstruction),
 		pendingChatAcquireInterval: pendingChatAcquireInterval,
+		maxChatsPerAcquire:         maxChatsPerAcquire,
 		inFlightChatStaleAfter:     inFlightChatStaleAfter,
 	}
 
@@ -1272,7 +1280,7 @@ func (p *Server) processOnce(ctx context.Context) {
 	chats, err := p.db.AcquireChats(acquireCtx, database.AcquireChatsParams{
 		StartedAt: time.Now(),
 		WorkerID:  p.workerID,
-		NumChats:  maxChatsPerAcquire,
+		NumChats:  p.maxChatsPerAcquire,
 	})
 	acquireCancel()
 	if err != nil {

coderd/coderd.go

@@ -10,6 +10,7 @@ import (
 	"flag"
 	"fmt"
 	"io"
+	"math"
 	"net/http"
 	httppprof "net/http/pprof"
 	"net/url"
@@ -766,17 +767,26 @@ func New(options *Options) *API {
 	}
 	api.agentProvider = stn
 
+	maxChatsPerAcquire := options.DeploymentValues.AI.Chat.AcquireBatchSize.Value()
+	if maxChatsPerAcquire > math.MaxInt32 {
+		maxChatsPerAcquire = math.MaxInt32
+	}
+	if maxChatsPerAcquire < math.MinInt32 {
+		maxChatsPerAcquire = math.MinInt32
+	}
+
 	api.chatDaemon = chatd.New(chatd.Config{
-		Logger:            options.Logger.Named("chats"),
-		Database:          options.Database,
-		ReplicaID:         api.ID,
-		SubscribeFn:       options.ChatSubscribeFn,
-		ProviderAPIKeys:   chatProviderAPIKeysFromDeploymentValues(options.DeploymentValues),
-		AgentConn:         api.agentProvider.AgentConn,
-		CreateWorkspace:   api.chatCreateWorkspace,
-		StartWorkspace:    api.chatStartWorkspace,
-		Pubsub:            options.Pubsub,
-		WebpushDispatcher: options.WebPushDispatcher,
+		Logger:             options.Logger.Named("chats"),
+		Database:           options.Database,
+		ReplicaID:          api.ID,
+		SubscribeFn:        options.ChatSubscribeFn,
+		MaxChatsPerAcquire: int32(maxChatsPerAcquire), //nolint:gosec // maxChatsPerAcquire is clamped to int32 range above.
+		ProviderAPIKeys:    chatProviderAPIKeysFromDeploymentValues(options.DeploymentValues),
+		AgentConn:          api.agentProvider.AgentConn,
+		CreateWorkspace:    api.chatCreateWorkspace,
+		StartWorkspace:     api.chatStartWorkspace,
+		Pubsub:             options.Pubsub,
+		WebpushDispatcher:  options.WebPushDispatcher,
 	})
 	gitSyncLogger := options.Logger.Named("gitsync")
 	refresher := gitsync.NewRefresher(