@juspay/neurolink 9.12.3 → 9.14.0

package/CHANGELOG.md

@@ -1,3 +1,15 @@
+## [9.14.0](https://github.com/juspay/neurolink/compare/v9.13.0...v9.14.0) (2026-02-27)
+
+### Features
+
+- **(docs):** documentation site overhaul — MCP docs server, homepage redesign, and SEO improvements ([5b16ed4](https://github.com/juspay/neurolink/commit/5b16ed4d5455568a480aa0389ad934eed9d03360))
+
+## [9.13.0](https://github.com/juspay/neurolink/compare/v9.12.3...v9.13.0) (2026-02-25)
+
+### Features
+
+- **(memory):** integrate Hippocampus SDK for enhanced user memory management ([4da4e63](https://github.com/juspay/neurolink/commit/4da4e635cb5175c10c2efc63643d97cdee301e25))
+
 ## [9.12.3](https://github.com/juspay/neurolink/compare/v9.12.2...v9.12.3) (2026-02-24)
 
 ### Bug Fixes

package/README.md

@@ -37,6 +37,7 @@ Extracted from production systems at Juspay and battle-tested at enterprise scal
 
 | Feature                             | Version | Description                                                                                                                                                   | Guide                                                                 |
 | ----------------------------------- | ------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------- |
+| **Memory**                          | v9.12.0 | Per-user condensed memory that persists across conversations. LLM-powered condensation with S3, Redis, or SQLite backends.                                    | [Memory Guide](docs/features/memory.md)                               |
 | **Context Window Management**       | v9.2.0  | 4-stage compaction pipeline with auto-detection, budget gate at 80% usage, per-provider token estimation                                                      | [Context Compaction Guide](docs/features/context-compaction.md)       |
 | **Tool Execution Control**          | v9.3.0  | `prepareStep` and `toolChoice` support for per-step tool enforcement in multi-step agentic loops. API-level control over tool calls.                          | [API Reference](docs/api/type-aliases/GenerateOptions.md#preparestep) |
 | **File Processor System**           | v9.1.0  | 17+ file type processors with ProcessorRegistry, security sanitization, SVG text injection                                                                    | [File Processors Guide](docs/features/file-processors.md)             |
@@ -48,6 +49,7 @@ Extracted from production systems at Juspay and battle-tested at enterprise scal
 | **Image Generation with Gemini**    | v8.31.0 | Native image generation using Gemini 2.0 Flash Experimental (`imagen-3.0-generate-002`). High-quality image synthesis directly from Google AI.                | [Image Generation Guide](docs/image-generation-streaming.md)          |
 | **HTTP/Streamable HTTP Transport**  | v8.29.0 | Connect to remote MCP servers via HTTP with authentication headers, automatic retry with exponential backoff, and configurable rate limiting.                 | [HTTP Transport Guide](docs/mcp-http-transport.md)                    |
 
+- **Memory** – Per-user condensed memory that persists across all conversations. Automatically retrieves and stores memory on each `generate()`/`stream()` call. Supports S3, Redis, and SQLite storage with LLM-powered condensation. → [Memory Guide](docs/features/memory.md)
 - **External TracerProvider Support** – Integrate NeuroLink with applications that already have OpenTelemetry instrumentation. Supports auto-detection and manual configuration. → [Observability Guide](docs/features/observability.md)
 - **Server Adapters** – Deploy NeuroLink as an HTTP API server with your framework of choice (Hono, Express, Fastify, Koa). Full CLI support with `serve` and `server` commands for foreground/background modes, route management, and OpenAPI generation. → [Server Adapters Guide](docs/guides/server-adapters/index.md)
 - **Title Generation Events** – Emit real-time events when conversation titles are auto-generated. Listen to `conversation:titleGenerated` for session tracking. → [Conversation Memory Guide](docs/conversation-memory.md#title-generation-events)
@@ -240,7 +242,7 @@ const result = await neurolink.generate({
 | --------------------------- | --------------------------------------------------------------------------------- | --------------------------------------------------------- |
 | **Auto Provider Selection** | Intelligent provider fallback                                                     | [SDK Guide](docs/sdk/index.md#auto-selection)             |
 | **Streaming Responses**     | Real-time token streaming                                                         | [Streaming Guide](docs/advanced/streaming.md)             |
-| **Conversation Memory**     | Automatic context management                                                      | [Memory Guide](docs/sdk/index.md#memory)                  |
+| **Conversation Memory**     | Automatic context management with embedded per-user memory                        | [Memory Guide](docs/sdk/index.md#memory)                  |
 | **Full Type Safety**        | Complete TypeScript types                                                         | [Type Reference](docs/sdk/api-reference.md)               |
 | **Error Handling**          | Graceful provider fallback                                                        | [Error Guide](docs/reference/troubleshooting.md)          |
 | **Analytics & Evaluation**  | Usage tracking, quality scores                                                    | [Analytics Guide](docs/advanced/analytics.md)             |
@@ -297,16 +299,17 @@ const result = await neurolink.generate({
 
 **Production-ready capabilities for regulated industries:**
 
-| Feature                     | Description                        | Use Case                  | Documentation                                               |
-| --------------------------- | ---------------------------------- | ------------------------- | ----------------------------------------------------------- |
-| **Enterprise Proxy**        | Corporate proxy support            | Behind firewalls          | [Proxy Setup](docs/enterprise-proxy-setup.md)               |
-| **Redis Memory**            | Distributed conversation state     | Multi-instance deployment | [Redis Guide](docs/getting-started/provider-setup.md#redis) |
-| **Cost Optimization**       | Automatic cheapest model selection | Budget control            | [Cost Guide](docs/advanced/index.md)                        |
-| **Multi-Provider Failover** | Automatic provider switching       | High availability         | [Failover Guide](docs/advanced/index.md)                    |
-| **Telemetry & Monitoring**  | OpenTelemetry integration          | Observability             | [Telemetry Guide](docs/telemetry-guide.md)                  |
-| **Security Hardening**      | Credential management, auditing    | Compliance                | [Security Guide](docs/advanced/enterprise.md)               |
-| **Custom Model Hosting**    | SageMaker integration              | Private models            | [SageMaker Guide](docs/sagemaker-integration.md)            |
-| **Load Balancing**          | LiteLLM proxy integration          | Scale & routing           | [Load Balancing](docs/litellm-integration.md)               |
+| Feature                     | Description                                 | Use Case                  | Documentation                                               |
+| --------------------------- | ------------------------------------------- | ------------------------- | ----------------------------------------------------------- |
+| **Enterprise Proxy**        | Corporate proxy support                     | Behind firewalls          | [Proxy Setup](docs/enterprise-proxy-setup.md)               |
+| **Redis Memory**            | Distributed conversation state              | Multi-instance deployment | [Redis Guide](docs/getting-started/provider-setup.md#redis) |
+| **Memory**                  | Per-user condensed memory (S3/Redis/SQLite) | Long-term user context    | [Memory Guide](docs/features/memory.md)                     |
+| **Cost Optimization**       | Automatic cheapest model selection          | Budget control            | [Cost Guide](docs/advanced/index.md)                        |
+| **Multi-Provider Failover** | Automatic provider switching                | High availability         | [Failover Guide](docs/advanced/index.md)                    |
+| **Telemetry & Monitoring**  | OpenTelemetry integration                   | Observability             | [Telemetry Guide](docs/telemetry-guide.md)                  |
+| **Security Hardening**      | Credential management, auditing             | Compliance                | [Security Guide](docs/advanced/enterprise.md)               |
+| **Custom Model Hosting**    | SageMaker integration                       | Private models            | [SageMaker Guide](docs/sagemaker-integration.md)            |
+| **Load Balancing**          | LiteLLM proxy integration                   | Scale & routing           | [Load Balancing](docs/litellm-integration.md)               |
 
 **Security & Compliance:**
 

package/dist/cli/commands/docs.d.ts

@@ -0,0 +1,10 @@
+import type { CommandModule } from "yargs";
+type DocsCommandArgs = {
+    transport?: "stdio" | "http";
+    port?: number;
+};
+export declare class DocsCommandFactory {
+    static createDocsCommand(): CommandModule<object, DocsCommandArgs>;
+    private static executeDocs;
+}
+export {};

package/dist/cli/commands/docs.js

@@ -0,0 +1,48 @@
+import chalk from "chalk";
+export class DocsCommandFactory {
+    static createDocsCommand() {
+        return {
+            command: "docs",
+            describe: "Start the NeuroLink documentation MCP server",
+            builder: (yargs) => yargs
+                .option("transport", {
+                alias: "t",
+                type: "string",
+                choices: ["stdio", "http"],
+                default: "stdio",
+                description: "Transport protocol (stdio for local, http for remote)",
+            })
+                .option("port", {
+                alias: "p",
+                type: "number",
+                default: 3001,
+                description: "Port for HTTP transport (ignored for stdio)",
+            }),
+            handler: async (argv) => {
+                await DocsCommandFactory.executeDocs(argv);
+            },
+        };
+    }
+    static async executeDocs(argv) {
+        try {
+            // Dynamic path prevents TypeScript from resolving outside rootDir
+            const mcpServerPath = "../../../docs-site/mcp-server/index.js";
+            const { startDocsServer } = await import(mcpServerPath);
+            await startDocsServer({
+                transport: argv.transport,
+                port: argv.port,
+            });
+        }
+        catch (err) {
+            if (err instanceof Error &&
+                err.message.includes("search-index.json not found")) {
+                console.error(chalk.red("\nSearch index not found. Build the docs site first:\n"));
+                console.error(chalk.cyan("  cd docs-site && pnpm build\n"));
+                process.exit(1);
+            }
+            console.error(chalk.red("Failed to start docs server:"), err);
+            process.exit(1);
+        }
+    }
+}
+//# sourceMappingURL=docs.js.map

package/dist/cli/parser.js

@@ -10,6 +10,7 @@ import { SetupCommandFactory } from "./factories/setupCommandFactory.js";
 import { ServerCommandFactory } from "./commands/server.js";
 import { ServeCommandFactory } from "./commands/serve.js";
 import { ragCommand } from "./commands/rag.js";
+import { DocsCommandFactory } from "./commands/docs.js";
 // Enhanced CLI with Professional UX
 export function initializeCliParser() {
     return (yargs(hideBin(process.argv))
@@ -166,6 +167,8 @@ export function initializeCliParser() {
         // Serve Command - Simplified server start - Using ServeCommandFactory
         .command(ServeCommandFactory.createServeCommands())
         // RAG Document Processing Commands
-        .command(ragCommand)); // Close the main return statement
+        .command(ragCommand)
+        // Docs MCP Server Command
+        .command(DocsCommandFactory.createDocsCommand())); // Close the main return statement
 }
 //# sourceMappingURL=parser.js.map

package/dist/index.d.ts

@@ -1,7 +1,7 @@
 /**
  * NeuroLink AI Toolkit
  *
- * A unified AI provider interface with support for 13+ providers,
+ * A unified AI provider interface with support for 14+ providers,
  * automatic fallback, streaming, MCP tool integration, HITL security,
  * Redis persistence, and enterprise-grade middleware.
  *
@@ -60,9 +60,9 @@ export declare const VERSION = "1.0.0";
  * Quick start factory function for creating AI provider instances.
  *
  * Creates a configured AI provider instance ready for immediate use.
- * Supports all 13 providers: OpenAI, Anthropic, Google AI Studio,
+ * Supports 14+ providers: OpenAI, Anthropic, Google AI Studio,
  * Google Vertex, AWS Bedrock, AWS SageMaker, Azure OpenAI, Hugging Face,
- * LiteLLM, Mistral, Ollama, OpenAI Compatible, and OpenRouter.
+ * LiteLLM, Mistral, Ollama, OpenAI Compatible, OpenRouter, and more.
  *
  * @category Factory
  *

package/dist/index.js

@@ -1,7 +1,7 @@
 /**
  * NeuroLink AI Toolkit
  *
- * A unified AI provider interface with support for 13+ providers,
+ * A unified AI provider interface with support for 14+ providers,
  * automatic fallback, streaming, MCP tool integration, HITL security,
  * Redis persistence, and enterprise-grade middleware.
  *
@@ -76,9 +76,9 @@ export const VERSION = "1.0.0";
  * Quick start factory function for creating AI provider instances.
  *
  * Creates a configured AI provider instance ready for immediate use.
- * Supports all 13 providers: OpenAI, Anthropic, Google AI Studio,
+ * Supports 14+ providers: OpenAI, Anthropic, Google AI Studio,
  * Google Vertex, AWS Bedrock, AWS SageMaker, Azure OpenAI, Hugging Face,
- * LiteLLM, Mistral, Ollama, OpenAI Compatible, and OpenRouter.
+ * LiteLLM, Mistral, Ollama, OpenAI Compatible, OpenRouter, and more.
  *
  * @category Factory
  *

package/dist/lib/index.d.ts

@@ -1,7 +1,7 @@
 /**
  * NeuroLink AI Toolkit
  *
- * A unified AI provider interface with support for 13+ providers,
+ * A unified AI provider interface with support for 14+ providers,
  * automatic fallback, streaming, MCP tool integration, HITL security,
  * Redis persistence, and enterprise-grade middleware.
  *
@@ -60,9 +60,9 @@ export declare const VERSION = "1.0.0";
  * Quick start factory function for creating AI provider instances.
  *
  * Creates a configured AI provider instance ready for immediate use.
- * Supports all 13 providers: OpenAI, Anthropic, Google AI Studio,
+ * Supports 14+ providers: OpenAI, Anthropic, Google AI Studio,
  * Google Vertex, AWS Bedrock, AWS SageMaker, Azure OpenAI, Hugging Face,
- * LiteLLM, Mistral, Ollama, OpenAI Compatible, and OpenRouter.
+ * LiteLLM, Mistral, Ollama, OpenAI Compatible, OpenRouter, and more.
  *
  * @category Factory
  *

package/dist/lib/index.js

@@ -1,7 +1,7 @@
 /**
  * NeuroLink AI Toolkit
  *
- * A unified AI provider interface with support for 13+ providers,
+ * A unified AI provider interface with support for 14+ providers,
  * automatic fallback, streaming, MCP tool integration, HITL security,
  * Redis persistence, and enterprise-grade middleware.
  *
@@ -76,9 +76,9 @@ export const VERSION = "1.0.0";
  * Quick start factory function for creating AI provider instances.
  *
  * Creates a configured AI provider instance ready for immediate use.
- * Supports all 13 providers: OpenAI, Anthropic, Google AI Studio,
+ * Supports 14+ providers: OpenAI, Anthropic, Google AI Studio,
  * Google Vertex, AWS Bedrock, AWS SageMaker, Azure OpenAI, Hugging Face,
- * LiteLLM, Mistral, Ollama, OpenAI Compatible, and OpenRouter.
+ * LiteLLM, Mistral, Ollama, OpenAI Compatible, OpenRouter, and more.
  *
  * @category Factory
  *

package/dist/lib/memory/hippocampusInitializer.d.ts

@@ -0,0 +1,6 @@
+import { Hippocampus, type HippocampusConfig } from "@juspay/hippocampus";
+export type { HippocampusConfig };
+export type Memory = HippocampusConfig & {
+    enabled?: boolean;
+};
+export declare function initializeHippocampus(config: HippocampusConfig): Hippocampus | null;

package/dist/lib/memory/hippocampusInitializer.js

@@ -0,0 +1,20 @@
+import { Hippocampus } from "@juspay/hippocampus";
+import { logger } from "../utils/logger.js";
+export function initializeHippocampus(config) {
+    try {
+        const instance = new Hippocampus(config);
+        logger.info("[memoryInitializer] Memory initialized successfully", {
+            storageType: config.storage?.type || "sqlite",
+            maxWords: config.maxWords || 50,
+            hasCustomPrompt: !!config.prompt,
+        });
+        return instance;
+    }
+    catch (error) {
+        logger.warn("[memoryInitializer] Failed to initialize memory; disabling", {
+            error: error instanceof Error ? error.message : String(error),
+        });
+        return null;
+    }
+}
+//# sourceMappingURL=hippocampusInitializer.js.map

package/dist/lib/neurolink.d.ts

@@ -24,7 +24,7 @@ import type { BatchOperationResult } from "./types/typeAliases.js";
 /**
  * NeuroLink - Universal AI Development Platform
  *
- * Main SDK class providing unified access to 13+ AI providers with enterprise features:
+ * Main SDK class providing unified access to 14+ AI providers with enterprise features:
  * - Multi-provider support (OpenAI, Anthropic, Google AI Studio, Google Vertex, AWS Bedrock, etc.)
  * - MCP (Model Context Protocol) tool integration with 58+ external servers
  * - Human-in-the-Loop (HITL) security workflows for regulated industries
@@ -126,6 +126,8 @@ export declare class NeuroLink {
     private _sessionCostUsd;
     private fileRegistry;
     private cachedFileTools;
+    private memoryInstance?;
+    private memorySDKConfig?;
     /**
      * Extract and set Langfuse context from options with proper async scoping
      */
@@ -138,6 +140,11 @@ export declare class NeuroLink {
      * Async initialization called during generate/stream
      */
     private ensureMem0Ready;
+    private initializeMemoryConfig;
+    /**
+     * Lazy initialization for memory — called during generate/stream.
+     */
+    private ensureMemoryReady;
     /**
      * Context storage for tool execution
      * This context will be merged with any runtime context passed by the AI model
@@ -228,6 +235,16 @@ export declare class NeuroLink {
     private extractMemoryContext;
     /** Store conversation turn in mem0 */
     private storeMem0ConversationTurn;
+    /**
+     * Retrieve condensed memory for a user.
+     * Returns the input text enhanced with memory context, or unchanged if no memory.
+     */
+    private retrieveMemory;
+    /**
+     * Store a conversation turn in memory (non-blocking).
+     * Calls add(userId, content) which internally condenses old + new via LLM.
+     */
+    private storeMemoryInBackground;
     /**
      * Set up HITL event forwarding to main emitter
      */
@@ -375,7 +392,7 @@ export declare class NeuroLink {
      * Generate AI response with comprehensive feature support.
      *
      * Primary method for AI generation with support for all NeuroLink features:
-     * - Multi-provider support (13+ providers)
+     * - Multi-provider support (14+ providers)
      * - MCP tool integration
      * - Structured JSON output with Zod schemas
      * - Conversation memory (Redis or in-memory)

package/dist/lib/neurolink.js

@@ -34,6 +34,7 @@ import { directToolsServer } from "./mcp/servers/agent/directToolsServer.js";
 import { MCPToolRegistry } from "./mcp/toolRegistry.js";
 import { initializeMem0 } from "./memory/mem0Initializer.js";
 import { createMemoryRetrievalTools } from "./memory/memoryRetrievalTools.js";
+import { initializeHippocampus, } from "./memory/hippocampusInitializer.js";
 import { flushOpenTelemetry, getLangfuseHealthStatus, initializeOpenTelemetry, isOpenTelemetryInitialized, setLangfuseContext, shutdownOpenTelemetry, } from "./services/server/ai/observability/instrumentation.js";
 import { getConversationMessages, storeConversationTurn, } from "./utils/conversationMemory.js";
 // Enhanced error handling imports
@@ -56,7 +57,7 @@ import { runWorkflow } from "./workflow/core/workflowRunner.js";
 /**
  * NeuroLink - Universal AI Development Platform
  *
- * Main SDK class providing unified access to 13+ AI providers with enterprise features:
+ * Main SDK class providing unified access to 14+ AI providers with enterprise features:
  * - Multi-provider support (OpenAI, Anthropic, Google AI Studio, Google Vertex, AWS Bedrock, etc.)
  * - MCP (Model Context Protocol) tool integration with 58+ external servers
  * - Human-in-the-Loop (HITL) security workflows for regulated industries
@@ -180,6 +181,9 @@ export class NeuroLink {
     fileRegistry;
     // Cached file tools to avoid redundant createFileTools() calls per generate/stream
     cachedFileTools = null;
+    // Memory instance and config
+    memoryInstance;
+    memorySDKConfig;
     /**
      * Extract and set Langfuse context from options with proper async scoping
      */
@@ -274,6 +278,32 @@ export class NeuroLink {
         this.mem0Instance = await initializeMem0(this.mem0Config);
         return this.mem0Instance;
     }
+    initializeMemoryConfig() {
+        const memory = this.conversationMemoryConfig?.conversationMemory?.memory;
+        if (!memory?.enabled) {
+            return false;
+        }
+        this.memorySDKConfig = memory;
+        return true;
+    }
+    /**
+     * Lazy initialization for memory — called during generate/stream.
+     */
+    ensureMemoryReady() {
+        if (this.memoryInstance !== undefined) {
+            return this.memoryInstance;
+        }
+        if (!this.initializeMemoryConfig()) {
+            this.memoryInstance = null;
+            return null;
+        }
+        if (!this.memorySDKConfig) {
+            this.memoryInstance = null;
+            return null;
+        }
+        this.memoryInstance = initializeHippocampus(this.memorySDKConfig);
+        return this.memoryInstance;
+    }
     /**
      * Context storage for tool execution
      * This context will be merged with any runtime context passed by the AI model
@@ -682,6 +712,39 @@ Current user's request: ${currentInput}`;
             async_mode: true,
         });
     }
+    /**
+     * Retrieve condensed memory for a user.
+     * Returns the input text enhanced with memory context, or unchanged if no memory.
+     */
+    async retrieveMemory(inputText, userId) {
+        const client = this.ensureMemoryReady();
+        if (!client) {
+            return inputText;
+        }
+        const memory = await client.get(userId);
+        if (!memory) {
+            return inputText;
+        }
+        return this.formatMemoryContext(memory, inputText);
+    }
+    /**
+     * Store a conversation turn in memory (non-blocking).
+     * Calls add(userId, content) which internally condenses old + new via LLM.
+     */
+    storeMemoryInBackground(originalPrompt, responseContent, userId) {
+        setImmediate(async () => {
+            try {
+                const client = this.ensureMemoryReady();
+                if (client) {
+                    const content = `User: ${originalPrompt}\nAssistant: ${responseContent}`;
+                    await client.add(userId, content);
+                }
+            }
+            catch (error) {
+                logger.warn("Memory storage failed:", error);
+            }
+        });
+    }
     /**
      * Set up HITL event forwarding to main emitter
      */
@@ -1480,7 +1543,7 @@ Current user's request: ${currentInput}`;
      * Generate AI response with comprehensive feature support.
      *
      * Primary method for AI generation with support for all NeuroLink features:
-     * - Multi-provider support (13+ providers)
+     * - Multi-provider support (14+ providers)
      * - MCP tool integration
      * - Structured JSON output with Zod schemas
      * - Conversation memory (Redis or in-memory)
@@ -1635,6 +1698,17 @@ Current user's request: ${currentInput}`;
                     logger.warn("Mem0 memory retrieval failed:", error);
                 }
             }
+            // Memory retrieval
+            if (this.conversationMemoryConfig?.conversationMemory?.memory?.enabled &&
+                options.context?.userId) {
+                try {
+                    options.input.text = await this.retrieveMemory(options.input.text, options.context.userId);
+                    logger.debug("Memory retrieval successful");
+                }
+                catch (error) {
+                    logger.warn("Memory retrieval failed:", error);
+                }
+            }
             const startTime = Date.now();
             // Apply orchestration if enabled and no specific provider/model requested
             if (this.enableOrchestration && !options.provider && !options.model) {
@@ -1863,6 +1937,12 @@ Current user's request: ${currentInput}`;
                 }
             });
         }
+        // Memory storage
+        if (this.conversationMemoryConfig?.conversationMemory?.memory?.enabled &&
+            options.context?.userId &&
+            generateResult.content?.trim()) {
+            this.storeMemoryInBackground(originalPrompt ?? "", generateResult.content.trim(), options.context.userId);
+        }
     }
     /**
      * Handle PPT generation mode
@@ -3150,6 +3230,17 @@ Current user's request: ${currentInput}`;
                 logger.warn("Mem0 memory retrieval failed:", error);
             }
         }
+        // Memory retrieval
+        if (this.conversationMemoryConfig?.conversationMemory?.memory?.enabled &&
+            options.context?.userId) {
+            try {
+                options.input.text = await this.retrieveMemory(options.input.text, options.context.userId);
+                logger.debug("Memory retrieval successful");
+            }
+            catch (error) {
+                logger.warn("Memory retrieval failed:", error);
+            }
+        }
         // Apply orchestration if enabled and no specific provider/model requested
         if (this.enableOrchestration && !options.provider && !options.model) {
             try {
@@ -3405,6 +3496,11 @@ Current user's request: ${currentInput}`;
                 }
             });
         }
+        if (this.conversationMemoryConfig?.conversationMemory?.memory?.enabled &&
+            enhancedOptions.context?.userId &&
+            accumulatedContent?.trim()) {
+            this.storeMemoryInBackground(originalPrompt ?? "", accumulatedContent.trim(), enhancedOptions.context?.userId);
+        }
     }
     /**
      * Validate stream input with comprehensive error reporting

package/dist/lib/types/conversation.d.ts

@@ -33,6 +33,8 @@
  * - Current time (ISO): `new Date().toISOString()`
  */
 import type { Mem0Config } from "../memory/mem0Initializer.js";
+import type { Memory } from "../memory/hippocampusInitializer.js";
+export type { Memory };
 /**
  * Configuration for conversation memory feature
  */
@@ -53,6 +55,8 @@ export type ConversationMemoryConfig = {
     mem0Enabled?: boolean;
     /** Configuration for mem0 cloud API integration */
     mem0Config?: Mem0Config;
+    /** Memory SDK config (condensed key-value memory per user). Set enabled: true to activate. */
+    memory?: Memory;
     /** Redis configuration (optional) - overrides environment variables */
     redisConfig?: RedisStorageConfig;
     /** Context compaction configuration */

package/dist/memory/hippocampusInitializer.d.ts

@@ -0,0 +1,6 @@
+import { Hippocampus, type HippocampusConfig } from "@juspay/hippocampus";
+export type { HippocampusConfig };
+export type Memory = HippocampusConfig & {
+    enabled?: boolean;
+};
+export declare function initializeHippocampus(config: HippocampusConfig): Hippocampus | null;

package/dist/memory/hippocampusInitializer.js

@@ -0,0 +1,19 @@
+import { Hippocampus } from "@juspay/hippocampus";
+import { logger } from "../utils/logger.js";
+export function initializeHippocampus(config) {
+    try {
+        const instance = new Hippocampus(config);
+        logger.info("[memoryInitializer] Memory initialized successfully", {
+            storageType: config.storage?.type || "sqlite",
+            maxWords: config.maxWords || 50,
+            hasCustomPrompt: !!config.prompt,
+        });
+        return instance;
+    }
+    catch (error) {
+        logger.warn("[memoryInitializer] Failed to initialize memory; disabling", {
+            error: error instanceof Error ? error.message : String(error),
+        });
+        return null;
+    }
+}

package/dist/neurolink.d.ts

@@ -24,7 +24,7 @@ import type { BatchOperationResult } from "./types/typeAliases.js";
 /**
  * NeuroLink - Universal AI Development Platform
  *
- * Main SDK class providing unified access to 13+ AI providers with enterprise features:
+ * Main SDK class providing unified access to 14+ AI providers with enterprise features:
  * - Multi-provider support (OpenAI, Anthropic, Google AI Studio, Google Vertex, AWS Bedrock, etc.)
  * - MCP (Model Context Protocol) tool integration with 58+ external servers
  * - Human-in-the-Loop (HITL) security workflows for regulated industries
@@ -126,6 +126,8 @@ export declare class NeuroLink {
     private _sessionCostUsd;
     private fileRegistry;
     private cachedFileTools;
+    private memoryInstance?;
+    private memorySDKConfig?;
     /**
      * Extract and set Langfuse context from options with proper async scoping
      */
@@ -138,6 +140,11 @@ export declare class NeuroLink {
      * Async initialization called during generate/stream
      */
     private ensureMem0Ready;
+    private initializeMemoryConfig;
+    /**
+     * Lazy initialization for memory — called during generate/stream.
+     */
+    private ensureMemoryReady;
     /**
      * Context storage for tool execution
      * This context will be merged with any runtime context passed by the AI model
@@ -228,6 +235,16 @@ export declare class NeuroLink {
     private extractMemoryContext;
     /** Store conversation turn in mem0 */
     private storeMem0ConversationTurn;
+    /**
+     * Retrieve condensed memory for a user.
+     * Returns the input text enhanced with memory context, or unchanged if no memory.
+     */
+    private retrieveMemory;
+    /**
+     * Store a conversation turn in memory (non-blocking).
+     * Calls add(userId, content) which internally condenses old + new via LLM.
+     */
+    private storeMemoryInBackground;
     /**
      * Set up HITL event forwarding to main emitter
      */
@@ -375,7 +392,7 @@ export declare class NeuroLink {
      * Generate AI response with comprehensive feature support.
      *
      * Primary method for AI generation with support for all NeuroLink features:
-     * - Multi-provider support (13+ providers)
+     * - Multi-provider support (14+ providers)
      * - MCP tool integration
      * - Structured JSON output with Zod schemas
      * - Conversation memory (Redis or in-memory)