[EPIC-D][LOU-D34] AgentType is optional and deprecated

CHANGELOG.md

@@ -12,6 +12,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - `createAgent()` agents can pause for approval (LOU-D21): a `needsApproval` tool no longer fails the run with "requires approval but no approvalStore". The run pauses (`finishReason: 'awaiting-approval'`) in a per-agent `InMemoryApprovalStore` (or the new `approvalStore` option), and `agent.approvals.list()` / `agent.approvals.resolve({ id, approved, note? })` continue it, in its session if it paused in one. The `approve` option decides calls in code without pausing.
 
 ### Changed
+- `AgentType` is optional and deprecated (LOU-D34, not breaking): `AgentBuilder.build()` no longer requires `setType()` and `AgentConfig.agentType` is now optional (`createAgent()` and `specToAgent()` agents carry no type). `AgentType`, `setType()` and the `agent-types` registry/validators are marked `@deprecated`: they have no runtime effect and will be removed in the next minor; they stay exported for now. Drop your `setType(...)` calls.
 - BREAKING: `encrypt()` now generates a random salt per call instead of a hardcoded one. Ciphertext produced before this change cannot be decrypted with the new code and must be re-encrypted.
 - BREAKING: The mock repositories (`MockAgentRepository`, `MockSessionRepository`, `MockResultRepository`, `MockMemoryRepository`, `MockAttachmentRepository` — previously re-exported from `src/data/mocks.ts` via the package root/`./data` subpath; this package has no `createMockRepositories` factory) are no longer exported from the package root. Import them from `@loushy/build-ai-agent/testing` instead.
 

README.md

@@ -213,10 +213,9 @@ you need `maxSteps`, approval gates, checkpoints, tracing hooks, or a
 API — there is no `new AgentExecutor()`.
 
 ```typescript
-import { AgentBuilder, AgentExecutor, AgentType, resolveProvider } from '@loushy/build-ai-agent';
+import { AgentBuilder, AgentExecutor, resolveProvider } from '@loushy/build-ai-agent';
 
 const agent = AgentBuilder.create()
-  .setType(AgentType.SmartAssistant)
   .setName('Customer Support Agent')
   .setPrompt('You are a helpful customer support assistant.')
   .build();
@@ -280,7 +279,7 @@ Both files are generated (`npm run docs:llms`) and checked in CI.
 
 ### Agents
 
-Agents combine a **type** (`AgentType.SmartAssistant`, etc.), a **prompt**,
+Agents combine a **prompt**,
 **tools**, optional **flows**, and conversation **memory**. Build one with
 `createAgent()` for the common case, or `AgentBuilder` when you need full
 control over the resulting `AgentConfig`.
@@ -476,11 +475,10 @@ To wire one child `AgentConfig` as a tool yourself, use `createDelegateTool()`
 run the same way):
 
 ```typescript
-import { AgentExecutor, createDelegateTool, ToolRegistry, AgentType } from '@loushy/build-ai-agent';
+import { AgentExecutor, createDelegateTool, ToolRegistry } from '@loushy/build-ai-agent';
 
 const billingAgent = {
   name: 'Billing Agent',
-  agentType: AgentType.SmartAssistant,
   prompt: 'You answer billing questions and look up invoices.',
 };
 
@@ -498,7 +496,6 @@ registry.register(
 
 const supportAgent = {
   name: 'Support Agent',
-  agentType: AgentType.SmartAssistant,
   prompt: 'You help customers. Delegate billing questions to the billing agent.',
   tools: { delegate_billing_agent: { tool: 'delegate_billing_agent' } },
 };

docs/api-overview.md

@@ -15,7 +15,7 @@ npm run docs:build   # writes docs/api/index.html
 | `createAgent(config)`         | Zero-config `{ send(message) }` agent from a `model` string or provider (+ instructions, tools). |
 | `AgentBuilder`                | Fluent builder for an `AgentConfig` (`AgentBuilder.create().setName(...)...build()`). |
 | `AgentExecutor.execute(opts)` | Static executor: runs an agent (LLM + tool-calling loop) and resolves to an `ExecutionResult`. |
-| `AgentType`                   | Agent type enum (e.g. `AgentType.SmartAssistant`).                          |
+| `AgentType`                   | Deprecated, no runtime effect: agents need no type (removed next minor).    |
 | `resumeAfterApproval()`       | Resume an execution paused for human approval.                             |
 | `InMemoryApprovalStore`       | Process-local `ApprovalStore`; the default store of `createAgent()` agents. |
 | `SessionAwaitingApprovalError` | Thrown by `execute()` when its `sessionId` is paused on an approval (see [Durable execution](./durable-execution.md)). |

docs/quick-start.md

@@ -155,12 +155,10 @@ checkpoints, tracing, ...). `AgentExecutor` is a static API - there is no
 import {
   AgentBuilder,
   AgentExecutor,
-  AgentType,
   createMockProvider,
 } from '@loushy/build-ai-agent';
 
 const agent = AgentBuilder.create()
-  .setType(AgentType.SmartAssistant)
   .setName('Customer Support Agent')
   .setPrompt('You are a helpful customer support assistant.')
   .build();

llms-full.txt

@@ -211,10 +211,9 @@ you need `maxSteps`, approval gates, checkpoints, tracing hooks, or a
 API — there is no `new AgentExecutor()`.
 
 ```typescript
-import { AgentBuilder, AgentExecutor, AgentType, resolveProvider } from '@loushy/build-ai-agent';
+import { AgentBuilder, AgentExecutor, resolveProvider } from '@loushy/build-ai-agent';
 
 const agent = AgentBuilder.create()
-  .setType(AgentType.SmartAssistant)
   .setName('Customer Support Agent')
   .setPrompt('You are a helpful customer support assistant.')
   .build();
@@ -278,7 +277,7 @@ Both files are generated (`npm run docs:llms`) and checked in CI.
 
 ### Agents
 
-Agents combine a **type** (`AgentType.SmartAssistant`, etc.), a **prompt**,
+Agents combine a **prompt**,
 **tools**, optional **flows**, and conversation **memory**. Build one with
 `createAgent()` for the common case, or `AgentBuilder` when you need full
 control over the resulting `AgentConfig`.
@@ -474,11 +473,10 @@ To wire one child `AgentConfig` as a tool yourself, use `createDelegateTool()`
 run the same way):
 
 ```typescript
-import { AgentExecutor, createDelegateTool, ToolRegistry, AgentType } from '@loushy/build-ai-agent';
+import { AgentExecutor, createDelegateTool, ToolRegistry } from '@loushy/build-ai-agent';
 
 const billingAgent = {
   name: 'Billing Agent',
-  agentType: AgentType.SmartAssistant,
   prompt: 'You answer billing questions and look up invoices.',
 };
 
@@ -496,7 +494,6 @@ registry.register(
 
 const supportAgent = {
   name: 'Support Agent',
-  agentType: AgentType.SmartAssistant,
   prompt: 'You help customers. Delegate billing questions to the billing agent.',
   tools: { delegate_billing_agent: { tool: 'delegate_billing_agent' } },
 };
@@ -1128,12 +1125,10 @@ checkpoints, tracing, ...). `AgentExecutor` is a static API - there is no
 import {
   AgentBuilder,
   AgentExecutor,
-  AgentType,
   createMockProvider,
 } from '@loushy/build-ai-agent';
 
 const agent = AgentBuilder.create()
-  .setType(AgentType.SmartAssistant)
   .setName('Customer Support Agent')
   .setPrompt('You are a helpful customer support assistant.')
   .build();
@@ -1219,7 +1214,7 @@ npm run docs:build   # writes docs/api/index.html
 | `createAgent(config)`         | Zero-config `{ send(message) }` agent from a `model` string or provider (+ instructions, tools). |
 | `AgentBuilder`                | Fluent builder for an `AgentConfig` (`AgentBuilder.create().setName(...)...build()`). |
 | `AgentExecutor.execute(opts)` | Static executor: runs an agent (LLM + tool-calling loop) and resolves to an `ExecutionResult`. |
-| `AgentType`                   | Agent type enum (e.g. `AgentType.SmartAssistant`).                          |
+| `AgentType`                   | Deprecated, no runtime effect: agents need no type (removed next minor).    |
 | `resumeAfterApproval()`       | Resume an execution paused for human approval.                             |
 | `InMemoryApprovalStore`       | Process-local `ApprovalStore`; the default store of `createAgent()` agents. |
 | `SessionAwaitingApprovalError` | Thrown by `execute()` when its `sessionId` is paused on an approval (see [Durable execution](https://github.com/LinuxDevil/agent-sdk/blob/main/docs/durable-execution.md)). |

src/agent-types/registry.ts

@@ -3,78 +3,62 @@ import { AgentType, AgentTypeDescriptor } from '../types';
 /**
  * Agent types registry
  * Contains all available agent type descriptors
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
  */
 export const agentTypesRegistry: AgentTypeDescriptor[] = [
   {
     type: AgentType.SmartAssistant,
-    description: {
-      ar: "المساعدون الأذكياء هم **وكلاء متعددو الأغراض**. يمكنهم استخدام الأدوات، على سبيل المثال، للتحقق من التقويم الخاص بك أو حجز أحداث جديدة. يمكن استخدامها أيضًا للاستطلاعات (مع مهام أخرى) ولكن يجب تخصيصها للقيام بذلك على مستوى الموجه.",
-      en: "Smart assistants are **general-purpose agents**. They can use tools, for example, checking your calendar or booking new events. They can also be used for surveys (mixed with other tasks) but need to be fine-tuned for doing so at the prompt level."    
-    },
+    description: { en: 'General-purpose agent that can use tools.' },
     supportsUserFacingUI: true,
     requiredTabs: ['prompt', 'expectedResult'],
-    displayName: {
-      'ar': 'مساعد ذكي [محادثة]',
-      'en': 'Smart assistant [Chat]'
-    }
+    displayName: { en: 'Smart assistant [Chat]' },
   },
   {
     type: AgentType.SurveyAgent,
-    description: {
-      ar: "يتم استخدام وكلاء الاستطلاع لجمع المعلومات أو الآراء من المستخدمين. بناءً على الإجابات السابقة، يمكنهم **ضبط الأسئلة التالية ديناميكيًا**. يحفظ هؤلاء الوكلاء الإجابات لمزيد من المعالجة بالتنسيق المطلوب. يمكنهم استبدال أدوات مثل **النماذج، الاستطلاعات، نماذج التسجيل** وما إلى ذلك.",
-      en: "Survey agents are used to collect information or opinions from users. Based on previous answers, they can **dynamically adjust** the next questions. These agents save the answers for further processing in the desired format. They can replace tools like **Forms, Polls, Intake forms** etc."    
-    },
+    description: { en: 'Collects information from users and adapts follow-up questions.' },
     supportsUserFacingUI: true,
     requiredTabs: ['prompt', 'expectedResult'],
-    displayName: {
-      'ar': 'وكيل استطلاع [محادثة]',
-      'en': 'Survey agent [Chat]'
-    }
+    displayName: { en: 'Survey agent [Chat]' },
   },
   {
     type: AgentType.CommerceAgent,
-    description: {
-      ar: "يُستخدم الوكلاء التجاريون **لبيع المنتجات أو الخدمات**. يمكن استخدامها في **التجارة الإلكترونية**، **حجز الخدمات**، سيناريوهات **b2b/cpq**. يعملون على **كتالوج المنتجات** ويمكن استخدامها **للبيع الإضافي** أو **البيع المتقاطع** للمنتجات.",
-      en: "Commerce agents are used to **sell products or services**. They can be used in **e-commerce**, **service booking**, **b2b/cpq** scenarios. They operate on the **product catalog** and can be used to **upsell** or **cross-sell** products."            
-    },
+    description: { en: 'Sells products or services from a catalog.' },
     supportsUserFacingUI: true,
     requiredTabs: ['prompt', 'expectedResult'],
-    displayName: {
-      'ar': 'مساعد مبيعات [محادثة]',
-      'en': 'Sales assistant [Chat]'
-    }
+    displayName: { en: 'Sales assistant [Chat]' },
   },
   {
     type: AgentType.Flow,
-    description: {
-      ar: "تتيح لك الوكلاء المعتمدون على التدفق إنشاء **سيناريوهات معقدة**. يمكن استخدامها **لأتمتة العمليات** وتطوير التطبيقات التي يتم استدعاؤها بواسطة API أو وكلاء آخرين، باستخدام **اللغة الطبيعية**. يمكن استخدامها **لإنشاء أشجار القرار** أو **التكاملات**.",
-      en: "Flow-based agents let you create **complex scenarios**. They can be used to **automate processes** and develop apps that are called by API or other agents, using **natural language**. They can be used to **create decision trees** or **integrations**."    
-    },
+    description: { en: 'Runs multi-step scenarios called by an API or other agents.' },
     supportsUserFacingUI: true,
     requiredTabs: [],
-    displayName: {
-      'ar': 'تطبيق / سير عمل [API]',
-      'en': 'App / Workflow [API]'
-    }
-  }
+    displayName: { en: 'App / Workflow [API]' },
+  },
 ];
 
 /**
  * Get agent type descriptor by type
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
  */
 export function getAgentTypeDescriptor(type: AgentType): AgentTypeDescriptor | undefined {
   return agentTypesRegistry.find(descriptor => descriptor.type === type);
 }
 
 /**
  * Get all agent type descriptors
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
  */
 export function getAllAgentTypeDescriptors(): AgentTypeDescriptor[] {
   return [...agentTypesRegistry];
 }
 
 /**
  * Check if agent type is valid
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
  */
 export function isValidAgentType(type: string): type is AgentType {
   return agentTypesRegistry.some(descriptor => descriptor.type === type);

src/agent-types/validators.ts

@@ -6,6 +6,10 @@
 
 /**
  * Validate agent configuration
+ *
+ * `agentType` is optional; it is only checked when present.
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
  */
 export function validateAgentConfig(config: Partial<AgentConfig>): { valid: boolean; errors: string[] } {
   const errors: string[] = [];
@@ -16,11 +20,7 @@
     );
   }
 
-  if (!config.agentType) {
-    errors.push(
-      'Agent type is required. Example: AgentBuilder.create().setType(AgentType.SmartAssistant)...build()'
-    );
-  } else if (!Object.values(AgentType).includes(config.agentType)) {
+  if (config.agentType !== undefined && !Object.values(AgentType).includes(config.agentType)) {
     errors.push(`Invalid agent type: ${config.agentType}`);
   }
 
@@ -32,8 +32,10 @@
 
 /**
  * Validate agent tools configuration
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
  */
 export function validateAgentTools(tools: Record<string, any>): { valid: boolean; errors: string[] } {
   const errors: string[] = [];
 
   for (const [key, config] of Object.entries(tools)) {

src/core/AgentBuilder.test.ts

@@ -25,18 +25,24 @@ describe('AgentBuilder', () => {
     }).toThrow('Agent name is required');
   });
 
-  it('should throw error when type is missing', () => {
+  it('builds without setType() (agentType is optional and deprecated)', () => {
+    const agent = new AgentBuilder().setName('Test Agent').build();
+
+    expect(agent.name).toBe('Test Agent');
+    expect(agent.agentType).toBeUndefined();
+    expect('agentType' in agent).toBe(false);
+  });
+
+  it('still rejects an invalid agent type when one is provided', () => {
     expect(() => {
-      new AgentBuilder()
-        .setName('Test Agent')
-        .build();
-    }).toThrow('Agent type is required');
+      AgentBuilder.from({ name: 'Test Agent', agentType: 'bogus' as AgentType }).build();
+    }).toThrow('Invalid agent type: bogus');
   });
 
   it('names the missing field and shows a corrective snippet (LOU-H12)', () => {
     expect(() => {
       new AgentBuilder().build();
-    }).toThrow(/Agent name is required\. Example:.*Agent type is required\. Example:/s);
+    }).toThrow(/Agent name is required\. Example: AgentBuilder\.create\(\)\.setName\('my-agent'\)/);
   });
 
   it('should add tools correctly', () => {

src/core/AgentBuilder.ts

@@ -11,6 +11,9 @@
 
   /**
    * Set agent type
+   *
+   * @deprecated Has no runtime effect and will be removed in the next minor
+   * release. `build()` no longer requires a type; just drop this call.
    */
   public setType(type: AgentType): this {
     this.config.agentType = type;
@@ -106,7 +109,7 @@
   /**
    * Set expected result schema
    */
   public setExpectedResult(schema: any): this {
     this.config.expectedResult = schema;
     return this;
   }
@@ -122,7 +125,7 @@
   /**
    * Set events
    */
   public setEvents(events: any[]): this {
     this.config.events = events;
     return this;
   }
@@ -130,7 +133,7 @@
   /**
    * Set settings
    */
   public setSettings(settings: Record<string, any>): this {
     this.config.settings = settings;
     return this;
   }
@@ -138,7 +141,7 @@
   /**
    * Set metadata
    */
   public setMetadata(metadata: Record<string, any>): this {
     this.config.metadata = metadata;
     return this;
   }
@@ -152,7 +155,7 @@
     return {
       id: this.config.id || nanoid(),
       name: this.config.name!,
-      agentType: this.config.agentType!,
+      ...(this.config.agentType !== undefined && { agentType: this.config.agentType }),
       locale: this.config.locale || 'en',
       prompt: this.config.prompt,
       expectedResult: this.config.expectedResult,

src/createAgent.ts

@@ -16,7 +16,6 @@
  */
 
 import { AgentBuilder } from './core/AgentBuilder';
-import { AgentType } from './types';
 import { AgentExecutor, ExecuteOptions, ExecutionResult } from './execution/AgentExecutor';
 import type { AgentRun } from './execution/agentRun';
 import { LLMProvider } from './providers/llm';
@@ -290,7 +289,6 @@ export function createAgent(config: CreateAgentConfig = {}): SimpleAgent {
   const { toolRegistry, toolsConfig } = registerTools(config.tools ?? {});
 
   const builder = AgentBuilder.create()
-    .setType(AgentType.SmartAssistant)
     .setName(config.name || 'agent')
     .setPrompt(instructions)
     .setTools(toolsConfig);

src/data/models.ts

@@ -12,7 +12,7 @@ import { AgentConfig, AgentFlow } from '../types';
 export class Agent {
   id?: string;
   name: string;
-  agentType: string;
+  agentType?: string;
   locale: string;
   prompt?: string;
   expectedResult?: any;

src/deploy/specExecution.ts

@@ -9,7 +9,7 @@
  * resolving providers/tools in a way that works on its own platform.
  */
 import { AgentBuilder } from '../core/AgentBuilder';
-import { AgentType, ToolConfiguration, ToolDescriptor } from '../types';
+import { ToolConfiguration, ToolDescriptor } from '../types';
 import { ToolRegistry } from '../tools/ToolRegistry';
 import { ExecuteOptions } from '../execution/AgentExecutor';
 import { LLMProvider } from '../providers/llm';
@@ -37,7 +37,6 @@ export function prepareSpecExecution(spec: AgentSpec, resolvers: SpecResolvers):
   }
 
   const agent = AgentBuilder.create()
-    .setType(AgentType.SmartAssistant)
     .setName(spec.name)
     .setPrompt(spec.prompt)
     .setTools(toolsConfig)

src/types/agent.ts

@@ -4,6 +4,9 @@ import { AgentFlow } from './flow';
 
 /**
  * Agent type identifiers
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
+ * Agents no longer need a type; omit `setType()` / `agentType`.
  */
 export enum AgentType {
   SmartAssistant = 'smart-assistant',
@@ -14,6 +17,8 @@ export enum AgentType {
 
 /**
  * Agent type descriptor
+ *
+ * @deprecated Has no runtime effect and will be removed in the next minor release.
  */
 export interface AgentTypeDescriptor {
   type: AgentType;
@@ -29,7 +34,8 @@ export interface AgentTypeDescriptor {
 export interface AgentConfig {
   id?: string;
   name: string;
-  agentType: AgentType;
+  /** @deprecated Optional and ignored at runtime; will be removed in the next minor release. */
+  agentType?: AgentType;
   locale?: string;
   prompt?: string;
   expectedResult?: any;