feat: Abort Signal Pass-Through Providers (Group A) — Issue #616

src/api/providers/__tests__/deepseek.spec.ts

@@ -494,7 +494,7 @@ describe("DeepSeekHandler", () => {
 				expect.objectContaining({
 					thinking: { type: "enabled" },
 				}),
-				{}, // Empty path options for non-Azure URLs
+				undefined, // No signal, non-Azure URL
 			)
 			const callArgs = mockCreate.mock.calls[0][0]
 			expect(callArgs.reasoning_effort).toBeUndefined()
@@ -517,7 +517,7 @@ describe("DeepSeekHandler", () => {
 					reasoning_effort: "high",
 					max_completion_tokens: 200_000,
 				}),
-				{},
+				undefined,
 			)
 		})
 
@@ -554,10 +554,27 @@ describe("DeepSeekHandler", () => {
 					thinking: { type: "enabled" },
 					reasoning_effort: "max",
 				}),
-				{},
+				undefined,
 			)
 		})
 
+		it("should pass Azure path even when abortSignal is not provided", async () => {
+			const azureHandler = new DeepSeekHandler({
+				...mockOptions,
+				deepSeekBaseUrl: "https://example.services.ai.azure.com",
+			})
+
+			const stream = azureHandler.createMessage(systemPrompt, messages)
+			for await (const _chunk of stream) {
+				// Consume the stream
+			}
+
+			expect(mockCreate).toHaveBeenCalledTimes(1)
+			const [requestOptions, signalArg] = mockCreate.mock.calls[0]
+			expect(requestOptions.model).toBe("deepseek-chat")
+			expect(signalArg).toBeUndefined()
+		})
+
 		it("should disable thinking for deepseek-v4 models when reasoning is disabled", async () => {
 			const v4Handler = new DeepSeekHandler({
 				...mockOptions,

src/api/providers/__tests__/lmstudio-native-tools.spec.ts

@@ -83,6 +83,7 @@ describe("LmStudioHandler Native Tools", () => {
 						}),
 					]),
 				}),
+				undefined,
 			)
 			// parallel_tool_calls should be true by default when not explicitly set
 			const callArgs = mockCreate.mock.calls[0][0]
@@ -109,6 +110,7 @@ describe("LmStudioHandler Native Tools", () => {
 				expect.objectContaining({
 					tool_choice: "auto",
 				}),
+				undefined,
 			)
 		})
 
@@ -221,9 +223,9 @@ describe("LmStudioHandler Native Tools", () => {
 				expect.objectContaining({
 					parallel_tool_calls: true,
 				}),
+				undefined,
 			)
 		})
-
 		it("should yield tool_call_end events when finish_reason is tool_calls", async () => {
 			mockCreate.mockImplementationOnce(() => ({
 				[Symbol.asyncIterator]: async function* () {

src/api/providers/__tests__/lmstudio.spec.ts

@@ -133,12 +133,15 @@ describe("LmStudioHandler", () => {
 		it("should complete prompt successfully", async () => {
 			const result = await handler.completePrompt("Test prompt")
 			expect(result).toBe("Test response")
-			expect(mockCreate).toHaveBeenCalledWith({
-				model: mockOptions.lmStudioModelId,
-				messages: [{ role: "user", content: "Test prompt" }],
-				temperature: 0,
-				stream: false,
-			})
+			expect(mockCreate).toHaveBeenCalledWith(
+				{
+					model: mockOptions.lmStudioModelId,
+					messages: [{ role: "user", content: "Test prompt" }],
+					temperature: 0,
+					stream: false,
+				},
+				undefined,
+			)
 		})
 
 		it("should handle API errors", async () => {

src/api/providers/__tests__/mimo.spec.ts

@@ -376,6 +376,7 @@ describe("MimoHandler", () => {
 				expect.objectContaining({
 					extra_body: { thinking: { type: "enabled" } },
 				}),
+				undefined,
 			)
 		})
 

src/api/providers/__tests__/minimax.spec.ts

@@ -305,6 +305,7 @@ describe("MiniMaxHandler", () => {
 					messages: expect.any(Array),
 					stream: true,
 				}),
+				undefined,
 			)
 		})
 
@@ -324,6 +325,7 @@ describe("MiniMaxHandler", () => {
 				expect.objectContaining({
 					temperature: 1,
 				}),
+				undefined,
 			)
 		})
 

src/api/providers/__tests__/requesty.spec.ts

@@ -210,6 +210,7 @@ describe("RequestyHandler", () => {
 					stream_options: { include_usage: true },
 					temperature: 0,
 				}),
+				undefined,
 			)
 		})
 
@@ -243,6 +244,7 @@ describe("RequestyHandler", () => {
 					thinking: { type: "adaptive" },
 					temperature: undefined,
 				}),
+				undefined,
 			)
 		})
 
@@ -314,6 +316,7 @@ describe("RequestyHandler", () => {
 						]),
 						tool_choice: "auto",
 					}),
+					undefined,
 				)
 			})
 
@@ -412,12 +415,15 @@ describe("RequestyHandler", () => {
 
 			expect(result).toBe("test completion")
 
-			expect(mockCreate).toHaveBeenCalledWith({
-				model: mockOptions.requestyModelId,
-				max_tokens: 8192,
-				messages: [{ role: "system", content: "test prompt" }],
-				temperature: 0,
-			})
+			expect(mockCreate).toHaveBeenCalledWith(
+				{
+					model: mockOptions.requestyModelId,
+					max_tokens: 8192,
+					messages: [{ role: "system", content: "test prompt" }],
+					temperature: 0,
+				},
+				undefined,
+			)
 		})
 
 		it("omits temperature for Claude Fable 5 in completePrompt", async () => {
@@ -429,12 +435,15 @@ describe("RequestyHandler", () => {
 
 			await handler.completePrompt("test prompt")
 
-			expect(mockCreate).toHaveBeenCalledWith({
-				model: "anthropic/claude-fable-5",
-				max_tokens: 8192,
-				messages: [{ role: "system", content: "test prompt" }],
-				temperature: undefined,
-			})
+			expect(mockCreate).toHaveBeenCalledWith(
+				{
+					model: "anthropic/claude-fable-5",
+					max_tokens: 8192,
+					messages: [{ role: "system", content: "test prompt" }],
+					temperature: undefined,
+				},
+				undefined,
+			)
 		})
 
 		it("handles API errors", async () => {

src/api/providers/__tests__/unbound.spec.ts

@@ -90,6 +90,7 @@ describe("UnboundHandler", () => {
 					mode: "architect",
 				},
 			}),
+			undefined,
 		)
 	})
 })

src/api/providers/__tests__/vercel-ai-gateway.spec.ts

@@ -263,9 +263,33 @@ describe("VercelAiGatewayHandler", () => {
 				expect.objectContaining({
 					temperature: customTemp,
 				}),
+				undefined,
 			)
 		})
 
+		it("forwards abortSignal to createMessage", async () => {
+			const controller = new AbortController()
+			const handler = new VercelAiGatewayHandler(mockOptions)
+
+			const systemPrompt = "You are a helpful assistant."
+			const messages: Anthropic.Messages.MessageParam[] = [{ role: "user", content: "Hello" }]
+
+			await handler.createMessage(systemPrompt, messages, { abortSignal: controller.signal }).next()
+
+			expect(mockCreate).toHaveBeenCalledWith(expect.objectContaining({}), { signal: controller.signal })
+		})
+
+		it("forwards abortSignal to completePrompt", async () => {
+			const controller = new AbortController()
+			const handler = new VercelAiGatewayHandler(mockOptions)
+
+			mockCreate.mockResolvedValue({ choices: [{ message: { content: "test completion" } }] })
+
+			await handler.completePrompt("test prompt", { abortSignal: controller.signal })
+
+			expect(mockCreate).toHaveBeenCalledWith(expect.objectContaining({}), { signal: controller.signal })
+		})
+
 		it("uses default temperature when none provided", async () => {
 			const handler = new VercelAiGatewayHandler(mockOptions)
 
@@ -278,6 +302,7 @@ describe("VercelAiGatewayHandler", () => {
 				expect.objectContaining({
 					temperature: VERCEL_AI_GATEWAY_DEFAULT_TEMPERATURE,
 				}),
+				undefined,
 			)
 		})
 
@@ -295,6 +320,7 @@ describe("VercelAiGatewayHandler", () => {
 					temperature: undefined,
 					max_completion_tokens: 128000,
 				}),
+				undefined,
 			)
 		})
 
@@ -325,6 +351,7 @@ describe("VercelAiGatewayHandler", () => {
 				expect.objectContaining({
 					max_completion_tokens: 64000, // max tokens for sonnet 4
 				}),
+				undefined,
 			)
 		})
 
@@ -403,6 +430,7 @@ describe("VercelAiGatewayHandler", () => {
 							}),
 						]),
 					}),
+					undefined,
 				)
 			})
 
@@ -420,6 +448,7 @@ describe("VercelAiGatewayHandler", () => {
 					expect.objectContaining({
 						tool_choice: "auto",
 					}),
+					undefined,
 				)
 			})
 
@@ -437,6 +466,7 @@ describe("VercelAiGatewayHandler", () => {
 					expect.objectContaining({
 						parallel_tool_calls: true,
 					}),
+					undefined,
 				)
 			})
 
@@ -454,6 +484,7 @@ describe("VercelAiGatewayHandler", () => {
 						tools: expect.any(Array),
 						parallel_tool_calls: true,
 					}),
+					undefined,
 				)
 			})
 
@@ -553,6 +584,7 @@ describe("VercelAiGatewayHandler", () => {
 					expect.objectContaining({
 						stream_options: { include_usage: true },
 					}),
+					undefined,
 				)
 			})
 		})
@@ -591,6 +623,7 @@ describe("VercelAiGatewayHandler", () => {
 					temperature: VERCEL_AI_GATEWAY_DEFAULT_TEMPERATURE,
 					max_completion_tokens: 64000,
 				}),
+				undefined,
 			)
 		})
 
@@ -607,6 +640,7 @@ describe("VercelAiGatewayHandler", () => {
 				expect.objectContaining({
 					temperature: customTemp,
 				}),
+				undefined,
 			)
 		})
 
@@ -655,6 +689,7 @@ describe("VercelAiGatewayHandler", () => {
 				expect.objectContaining({
 					temperature: 0.9,
 				}),
+				undefined,
 			)
 		})
 	})

src/api/providers/__tests__/zai.spec.ts

@@ -502,6 +502,7 @@ describe("ZAiHandler", () => {
 					model: "glm-5.1",
 					max_tokens: 40_000,
 				}),
+				undefined,
 			)
 		})
 
@@ -542,6 +543,7 @@ describe("ZAiHandler", () => {
 					model: "glm-5.1",
 					max_tokens: 100_000,
 				}),
+				undefined,
 			)
 		})
 
@@ -572,6 +574,7 @@ describe("ZAiHandler", () => {
 					model: "glm-4.7",
 					thinking: { type: "enabled" },
 				}),
+				undefined,
 			)
 		})
 
@@ -603,6 +606,7 @@ describe("ZAiHandler", () => {
 					model: "glm-4.7",
 					thinking: { type: "disabled" },
 				}),
+				undefined,
 			)
 		})
 
@@ -634,6 +638,7 @@ describe("ZAiHandler", () => {
 					model: "glm-4.7",
 					thinking: { type: "enabled" },
 				}),
+				undefined,
 			)
 		})
 
@@ -687,6 +692,7 @@ describe("ZAiHandler", () => {
 					model: "glm-5-turbo",
 					thinking: { type: "enabled" },
 				}),
+				undefined,
 			)
 		})
 
@@ -717,6 +723,7 @@ describe("ZAiHandler", () => {
 					model: "glm-5-turbo",
 					thinking: { type: "disabled" },
 				}),
+				undefined,
 			)
 		})
 	})

src/api/providers/deepseek.ts

@@ -133,7 +133,11 @@ export class DeepSeekHandler extends OpenAiHandler {
 		try {
 			stream = await this.client.chat.completions.create(
 				requestOptions as OpenAI.Chat.Completions.ChatCompletionCreateParamsStreaming,
-				isAzureAiInference ? { path: OPENAI_AZURE_AI_INFERENCE_PATH } : {},
+				metadata?.abortSignal
+					? isAzureAiInference
+						? { path: OPENAI_AZURE_AI_INFERENCE_PATH, signal: metadata.abortSignal }
+						: { signal: metadata.abortSignal }
+					: undefined,
 			)
 		} catch (error) {
 			const { handleOpenAIError } = await import("./utils/openai-error-handler")