diff --git a/packages/ai/CHANGELOG.md b/packages/ai/CHANGELOG.md
index 1051de9d..33314368 100644
--- a/packages/ai/CHANGELOG.md
+++ b/packages/ai/CHANGELOG.md
@@ -11,6 +11,7 @@
 - Fixed Amazon Bedrock prompt-cache points for Claude Fable 5 and Claude Sonnet 5 ([#6235](https://github.com/earendil-works/pi/issues/6235)).
 - Fixed DS4 server context overflow detection for `Prompt has ... tokens, but the configured context size is ... tokens` errors ([#6262](https://github.com/earendil-works/pi/issues/6262)).
 - Fixed OpenAI Codex WebSocket sessions to rotate cached connections before the backend's 60-minute limit, avoiding connection-limit failures on long sessions ([#6268](https://github.com/earendil-works/pi/issues/6268)).
+- Fixed OpenAI Completions and Responses providers to send `(no tool output)` instead of `(see attached image)` when a tool result has empty text and no image content, preventing the model from hallucinating image attachments.
 - Fixed OpenAI Responses and Azure OpenAI Responses requests to avoid sending `max_output_tokens` values below the provider minimum ([#6265](https://github.com/earendil-works/pi/issues/6265)).
 - Fixed retry classification for Cloudflare 524 timeout responses ([#6239](https://github.com/earendil-works/pi/issues/6239)).
 
diff --git a/packages/ai/src/api/openai-completions.ts b/packages/ai/src/api/openai-completions.ts
index b7cbb00e..c1d9f933 100644
--- a/packages/ai/src/api/openai-completions.ts
+++ b/packages/ai/src/api/openai-completions.ts
@@ -1033,10 +1033,11 @@ export function convertMessages(
 
 				// Always send tool result with text (or placeholder if only images)
 				const hasText = textResult.length > 0;
+				const toolResultText = hasText ? textResult : hasImages ? "(see attached image)" : "(no tool output)";
 				// Some providers require the 'name' field in tool results
 				const toolResultMsg: ChatCompletionToolMessageParam = {
 					role: "tool",
-					content: sanitizeSurrogates(hasText ? textResult : "(see attached image)"),
+					content: sanitizeSurrogates(toolResultText),
 					tool_call_id: toolMsg.toolCallId,
 				};
 				if (compat.requiresToolResultName && toolMsg.toolName) {
diff --git a/packages/ai/src/api/openai-responses-shared.ts b/packages/ai/src/api/openai-responses-shared.ts
index 9e58046e..527dea5c 100644
--- a/packages/ai/src/api/openai-responses-shared.ts
+++ b/packages/ai/src/api/openai-responses-shared.ts
@@ -251,7 +251,7 @@ export function convertResponsesMessages<TApi extends Api>(
 
 				output = contentParts;
 			} else {
-				output = sanitizeSurrogates(hasText ? textResult : "(see attached image)");
+				output = sanitizeSurrogates(hasText ? textResult : hasImages ? "(see attached image)" : "(no tool output)");
 			}
 
 			messages.push({
diff --git a/packages/ai/test/openai-completions-tool-result-images.test.ts b/packages/ai/test/openai-completions-tool-result-images.test.ts
index c8500792..31be1667 100644
--- a/packages/ai/test/openai-completions-tool-result-images.test.ts
+++ b/packages/ai/test/openai-completions-tool-result-images.test.ts
@@ -54,6 +54,17 @@ function buildToolResult(toolCallId: string, timestamp: number): ToolResultMessa
 	};
 }
 
+function buildEmptyToolResult(toolCallId: string, timestamp: number): ToolResultMessage {
+	return {
+		role: "toolResult",
+		toolCallId,
+		toolName: "bash",
+		content: [{ type: "text", text: "" }],
+		isError: false,
+		timestamp,
+	};
+}
+
 describe("openai-completions convertMessages", () => {
 	it("batches tool-result images after consecutive tool results", () => {
 		const { compat: _compat, ...baseModel } = getModel("openai", "gpt-4o-mini");
@@ -100,4 +111,39 @@ describe("openai-completions convertMessages", () => {
 		);
 		expect(imageParts.length).toBe(2);
 	});
+
+	it("uses '(no tool output)' placeholder for empty tool results without images", () => {
+		const { compat: _compat, ...baseModel } = getModel("openai", "gpt-4o-mini");
+		const model: Model<"openai-completions"> = {
+			...baseModel,
+			api: "openai-completions",
+			input: ["text", "image"],
+		};
+
+		const now = Date.now();
+		const assistantMessage: AssistantMessage = {
+			role: "assistant",
+			content: [{ type: "toolCall", id: "tool-1", name: "bash", arguments: { command: "true" } }],
+			api: model.api,
+			provider: model.provider,
+			model: model.id,
+			usage: emptyUsage,
+			stopReason: "toolUse",
+			timestamp: now,
+		};
+
+		const context: Context = {
+			messages: [
+				{ role: "user", content: "Run the command", timestamp: now - 1 },
+				assistantMessage,
+				buildEmptyToolResult("tool-1", now + 1),
+			],
+		};
+
+		const messages = convertMessages(model, context, compat);
+		const toolMessage = messages.find((m) => m.role === "tool") as { role: "tool"; content: string } | undefined;
+		expect(toolMessage).toBeTruthy();
+		expect(toolMessage?.content).toBe("(no tool output)");
+		expect(toolMessage?.content).not.toContain("see attached image");
+	});
 });
diff --git a/packages/ai/test/openai-responses-empty-tool-result.test.ts b/packages/ai/test/openai-responses-empty-tool-result.test.ts
new file mode 100644
index 00000000..0db91237
--- /dev/null
+++ b/packages/ai/test/openai-responses-empty-tool-result.test.ts
@@ -0,0 +1,58 @@
+import { describe, expect, it } from "vitest";
+import { convertResponsesMessages } from "../src/api/openai-responses-shared.ts";
+import { getModel } from "../src/compat.ts";
+import type { AssistantMessage, Context, ToolResultMessage, Usage } from "../src/types.ts";
+
+const usage: Usage = {
+	input: 0,
+	output: 0,
+	cacheRead: 0,
+	cacheWrite: 0,
+	totalTokens: 0,
+	cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0, total: 0 },
+};
+
+function buildEmptyToolResult(toolCallId: string, timestamp: number): ToolResultMessage {
+	return {
+		role: "toolResult",
+		toolCallId,
+		toolName: "bash",
+		content: [{ type: "text", text: "" }],
+		isError: false,
+		timestamp,
+	};
+}
+
+describe("OpenAI Responses convertResponsesMessages empty tool result", () => {
+	it("uses '(no tool output)' placeholder for empty tool results without images", () => {
+		const model = getModel("openai", "gpt-4o-mini");
+		const now = Date.now();
+		const assistant: AssistantMessage = {
+			role: "assistant",
+			content: [{ type: "toolCall", id: "tool-1", name: "bash", arguments: { command: "true" } }],
+			api: model.api,
+			provider: model.provider,
+			model: model.id,
+			usage,
+			stopReason: "toolUse",
+			timestamp: now,
+		};
+
+		const context: Context = {
+			messages: [
+				{ role: "user", content: "Run the command", timestamp: now - 1 },
+				assistant,
+				buildEmptyToolResult("tool-1", now + 1),
+			],
+		};
+
+		const input = convertResponsesMessages(model, context, new Set(["openai", "openai-codex", "opencode"]));
+		const functionCallOutput = input.find((item) => item.type === "function_call_output") as
+			| { type: "function_call_output"; output: string }
+			| undefined;
+
+		expect(functionCallOutput).toBeTruthy();
+		expect(functionCallOutput?.output).toBe("(no tool output)");
+		expect(functionCallOutput?.output).not.toContain("see attached image");
+	});
+});