fix(ai): keep global memory always enabled

Remove the proactive-memory preference and its cross-platform runtime plumbing so memory behavior has one clear contract. Normalize away the obsolete setting and keep source type limited to provenance.

apps/cli/src/ai/agent-stream-runner.ts

@@ -86,7 +86,6 @@ export function createCliRunAgentStream(
     crossChatAnalysisService?: CrossChatAnalysisService
     sessionAdapter?: SessionRuntimeAdapter
     memoryService?: AIMemoryService
-    getAllowProactiveMemory?: () => boolean
   } = {}
 ): (params: AgentStreamRequest, onEvent: (chunk: AgentStreamChunk) => void, abortSignal: AbortSignal) => Promise<void> {
   const aiDataDir = getAiDir(dbManager)
@@ -161,7 +160,6 @@ export function createCliRunAgentStream(
         maxToolResultTokens: resolveCrossChatToolResultTokenBudget(contextWindow),
         memoryService: options.memoryService,
         aiChatId,
-        allowProactiveMemory: options.getAllowProactiveMemory?.() ?? true,
       })
       await runCrossChatAgent({
         userMessage,

apps/cli/src/ai/cross-chat-tool-adapter.ts

@@ -23,7 +23,6 @@ export function createCliCrossChatTools(options: {
   maxToolResultTokens: number
   memoryService: AIMemoryService
   aiChatId: string
-  allowProactiveMemory: boolean
 }): AgentTool<any, any>[] {
   const context: Omit<CrossChatToolExecutionContext, 'abortSignal' | 'reportProgress'> = {
     locale: options.locale,
@@ -31,7 +30,6 @@ export function createCliCrossChatTools(options: {
     analysisService: options.analysisService,
     memoryService: options.memoryService,
     aiChatId: options.aiChatId,
-    allowProactiveMemory: options.allowProactiveMemory,
     maxToolResultTokens: options.maxToolResultTokens,
     countTokens,
     preprocessMessagesBySession: (sessionId, messages) =>

apps/cli/src/http/index.ts

@@ -233,7 +233,6 @@ export async function startHttpServer(options?: HttpServerOptions): Promise<{
         crossChatAnalysisService,
         sessionAdapter,
         memoryService: aiMemoryService,
-        getAllowProactiveMemory: () => preferencesManager.load().aiGlobalSettings.allowProactiveMemory,
       }),
     },
   })