Add image attachments to agent creation

packages/app/src/app/agent/new.tsx

@@ -573,12 +573,6 @@ export default function DraftAgentScreen() {
         ...(modeId ? { modeId } : {}),
         ...(trimmedModel ? { model: trimmedModel } : {}),
       };
-      // TODO: Images in initial agent creation are not yet supported by the server API.
-      // For now we log a warning. Images can be sent after agent creation via sendAgentMessage.
-      if (images && images.length > 0) {
-        console.warn("[DraftAgentScreen] Image attachments on agent creation not yet supported");
-      }
-
       const trimmedBaseBranch = baseBranch.trim();
       const shouldIncludeBase =
         trimmedBaseBranch.length > 0 ||
@@ -608,6 +602,7 @@ export default function DraftAgentScreen() {
       createAgent({
         config,
         initialPrompt: trimmedPrompt,
+        images,
         git: gitOptions,
         requestId,
       });

packages/app/src/contexts/session-context.tsx

@@ -1284,6 +1284,34 @@ export function SessionProvider({ children, serverUrl, serverId }: SessionProvid
     ws.send(msg);
   }, [serverId, ws, setProviderModels]);
 
+  const encodeImages = useCallback(async (
+    images?: Array<{ uri: string; mimeType?: string }>
+  ) => {
+    if (!images || images.length === 0) {
+      return undefined;
+    }
+    const encodedImages = await Promise.all(
+      images.map(async ({ uri, mimeType }) => {
+        try {
+          const data = await FileSystem.readAsStringAsync(uri, {
+            encoding: "base64",
+          });
+          return {
+            data,
+            mimeType: mimeType ?? "image/jpeg",
+          };
+        } catch (error) {
+          console.error("[Session] Failed to convert image:", error);
+          return null;
+        }
+      })
+    );
+    const validImages = encodedImages.filter(
+      (entry): entry is { data: string; mimeType: string } => entry !== null
+    );
+    return validImages.length > 0 ? validImages : undefined;
+  }, []);
+
   const sendAgentMessage = useCallback(async (
     agentId: string,
     message: string,
@@ -1304,31 +1332,7 @@ export function SessionProvider({ children, serverUrl, serverId }: SessionProvid
       return updated;
     });
 
-    let imagesData: Array<{ data: string; mimeType: string }> | undefined;
-    if (images && images.length > 0) {
-      const encodedImages = await Promise.all(
-        images.map(async ({ uri, mimeType }) => {
-          try {
-            const data = await FileSystem.readAsStringAsync(uri, {
-              encoding: "base64",
-            });
-            return {
-              data,
-              mimeType: mimeType ?? "image/jpeg",
-            };
-          } catch (error) {
-            console.error("[Session] Failed to convert image:", error);
-            return null;
-          }
-        })
-      );
-      const validImages = encodedImages.filter(
-        (entry): entry is { data: string; mimeType: string } => entry !== null
-      );
-      if (validImages.length > 0) {
-        imagesData = validImages;
-      }
-    }
+    const imagesData = await encodeImages(images);
 
     const msg: WSInboundMessage = {
       type: "session",
@@ -1341,7 +1345,7 @@ export function SessionProvider({ children, serverUrl, serverId }: SessionProvid
       },
     };
     ws.send(msg);
-  }, [serverId, ws, setAgentStreamState]);
+  }, [encodeImages, serverId, ws, setAgentStreamState]);
 
   const cancelAgentRun = useCallback((agentId: string) => {
     const msg: WSInboundMessage = {
@@ -1431,21 +1435,28 @@ export function SessionProvider({ children, serverUrl, serverId }: SessionProvid
     }
   }, [ws]);
 
-  const createAgent = useCallback(({ config, initialPrompt, git, worktreeName, requestId }: { config: any; initialPrompt: string; git?: any; worktreeName?: string; requestId?: string }) => {
+  const createAgent = useCallback(async ({ config, initialPrompt, images, git, worktreeName, requestId }: { config: any; initialPrompt: string; images?: Array<{ uri: string; mimeType?: string }>; git?: any; worktreeName?: string; requestId?: string }) => {
     const trimmedPrompt = initialPrompt.trim();
+    let imagesData: Array<{ data: string; mimeType: string }> | undefined;
+    try {
+      imagesData = await encodeImages(images);
+    } catch (error) {
+      console.error("[Session] Failed to prepare images for agent creation:", error);
+    }
     const msg: WSInboundMessage = {
       type: "session",
       message: {
         type: "create_agent_request",
         config,
         ...(trimmedPrompt ? { initialPrompt: trimmedPrompt } : {}),
+        ...(imagesData && imagesData.length > 0 ? { images: imagesData } : {}),
         ...(git ? { git } : {}),
         ...(worktreeName ? { worktreeName } : {}),
         ...(requestId ? { requestId } : {}),
       },
     };
     ws.send(msg);
-  }, [ws]);
+  }, [encodeImages, ws]);
 
   const resumeAgent = useCallback(({ handle, overrides, requestId }: { handle: any; overrides?: any; requestId?: string }) => {
     const msg: WSInboundMessage = {

packages/app/src/stores/session-store.ts

@@ -199,10 +199,11 @@ export interface SessionState {
     createAgent: (options: {
       config: any;
       initialPrompt: string;
+      images?: Array<{ uri: string; mimeType?: string }>;
       git?: any;
       worktreeName?: string;
       requestId?: string;
-    }) => void;
+    }) => Promise<void>;
     resumeAgent: (options: { handle: any; overrides?: any; requestId?: string }) => void;
     setAgentMode: (agentId: string, modeId: string) => void;
     respondToPermission: (agentId: string, requestId: string, response: any) => void;

packages/server/src/server/messages.ts

@@ -365,6 +365,10 @@ export const CreateAgentRequestMessageSchema = z.object({
   config: AgentSessionConfigSchema,
   worktreeName: z.string().optional(),
   initialPrompt: z.string().optional(),
+  images: z.array(z.object({
+    data: z.string(), // base64 encoded image
+    mimeType: z.string(), // e.g., "image/jpeg", "image/png"
+  })).optional(),
   git: GitSetupOptionsSchema.optional(),
   requestId: z.string().optional(),
 });

packages/server/src/server/session.ts

@@ -1265,7 +1265,7 @@ export class Session {
   private async handleCreateAgentRequest(
     msg: Extract<SessionInboundMessage, { type: "create_agent_request" }>
   ): Promise<void> {
-    const { config, worktreeName, requestId, initialPrompt, git } = msg;
+    const { config, worktreeName, requestId, initialPrompt, git, images } = msg;
     console.log(
       `[Session ${this.clientId}] Creating agent in ${config.cwd} (${config.provider})${
         worktreeName ? ` with worktree ${worktreeName}` : ""
@@ -1306,7 +1306,8 @@ export class Session {
           await this.handleSendAgentMessage(
             snapshot.id,
             trimmedPrompt,
-            uuidv4()
+            uuidv4(),
+            images
           );
         } catch (promptError) {
           console.error(

plan.md

@@ -134,10 +134,11 @@ Improvements to the new agent screen in the app.
     - If broken, add fix tasks with debugging details.
     - **Done (2025-12-29 09:13)**: WHAT: Tested image attachments via Playwright MCP; traced code in `packages/app/src/app/agent/new.tsx:576-580`, `packages/app/src/contexts/session-context.tsx:1307-1331`, `packages/server/src/server/session.ts:316-334`. RESULT: Image attachments are BROKEN - three issues found: (1) Agent creation silently drops images with console warning, (2) Web platform fails to base64 encode images due to `expo-file-system.readAsStringAsync` not working, (3) Server `buildAgentPrompt` only creates text summary, never passes actual image bytes to LLM. EVIDENCE: Console logs `[WARNING] [DraftAgentScreen] Image attachments on agent creation not yet supported` and `[ERROR] [Session] Failed to convert image: Error: Method readAsStringAsync imported from "expo-file-system"...`; agent responses consistently state "I don't see any image attached". Full report in `REPORT-image-attachment-investigation.md`.
 
-- [ ] **Fix**: Add image support to agent creation API
+- [x] **Fix**: Add image support to agent creation API
     - Extend `create_agent_request` to accept images array
     - Pass images to `handleSendAgentMessage` after agent creation
     - Remove warning in `new.tsx:576-580` once implemented
+    - **Done (2025-12-29 09:07)**: WHAT: `packages/server/src/server/messages.ts:363-373` adds optional `images` to `create_agent_request`; `packages/server/src/server/session.ts:1268-1311` forwards images into `handleSendAgentMessage`; `packages/app/src/contexts/session-context.tsx:1287-1459` encodes image attachments and includes them in create-agent websocket payloads; `packages/app/src/stores/session-store.ts:175-206` updates createAgent type to accept images/async; `packages/app/src/app/agent/new.tsx:560-606` forwards images to createAgent and removes the warning. RESULT: agent creation requests now carry image attachments through to the initial prompt path. EVIDENCE: Not run (not requested).
 
 - [ ] **Fix**: Fix web platform image base64 encoding
     - Replace `FileSystem.readAsStringAsync` in `session-context.tsx:1312` with cross-platform solution