From 2cb057fcd90fe1a0cbc1e63923846c705cee4820 Mon Sep 17 00:00:00 2001
From: Peter Steinberger <steipete@gmail.com>
Date: Mon, 6 Apr 2026 03:21:50 +0100
Subject: [PATCH] fix: harden async media completion delivery

---
 src/agents/tools/media-generate-background-shared.ts        | 5 +++--
 src/agents/tools/music-generate-background.test.ts          | 6 +++---
 src/agents/tools/video-generate-background.test.ts          | 6 +++---
 .../reply/dispatch-from-config.reply-dispatch.test.ts       | 4 ++--
 4 files changed, 11 insertions(+), 10 deletions(-)

diff --git a/src/agents/tools/media-generate-background-shared.ts b/src/agents/tools/media-generate-background-shared.ts
index b0a273694c0..959db3b78fe 100644
--- a/src/agents/tools/media-generate-background-shared.ts
+++ b/src/agents/tools/media-generate-background-shared.ts
@@ -1,5 +1,6 @@
 import crypto from "node:crypto";
 import { parseReplyDirectives } from "../../auto-reply/reply/reply-directives.js";
+import { SILENT_REPLY_TOKEN } from "../../auto-reply/tokens.js";
 import { createSubsystemLogger } from "../../logging/subsystem.js";
 import { parseAgentSessionKey } from "../../sessions/session-key-utils.js";
 import {
@@ -147,8 +148,8 @@ function buildMediaGenerationReplyInstruction(params: {
   if (params.status === "ok") {
     return [
       `A completed ${params.completionLabel} generation task is ready for user delivery.`,
-      `Reply in your normal assistant voice and post the finished ${params.completionLabel} to the original message channel now.`,
-      "If the result includes MEDIA: lines, include those exact MEDIA: lines in your reply so OpenClaw attaches the generated media.",
+      `Prefer the message tool for delivery: use action="send" to the current/original chat, put your user-facing caption in message, attach each generated file with path/filePath using the exact path from the result, then reply ONLY: ${SILENT_REPLY_TOKEN}.`,
+      `If you cannot use the message tool, reply in your normal assistant voice and include the exact MEDIA: lines from the result so OpenClaw attaches the finished ${params.completionLabel}.`,
       "Keep internal task/session details private and do not copy the internal event text verbatim.",
     ].join(" ");
   }
diff --git a/src/agents/tools/music-generate-background.test.ts b/src/agents/tools/music-generate-background.test.ts
index 6b9eb7ce5c5..3ce97c6630f 100644
--- a/src/agents/tools/music-generate-background.test.ts
+++ b/src/agents/tools/music-generate-background.test.ts
@@ -148,16 +148,16 @@ describe("music generate background helpers", () => {
           to: "channel:1",
         }),
         expectsCompletionMessage: true,
-        internalEvents: [
+        internalEvents: expect.arrayContaining([
           expect.objectContaining({
             source: "music_generation",
             announceType: "music generation task",
             status: "ok",
             result: expect.stringContaining("MEDIA:/tmp/generated-night-drive.mp3"),
             mediaUrls: ["/tmp/generated-night-drive.mp3"],
-            replyInstruction: expect.stringContaining("include those exact MEDIA: lines"),
+            replyInstruction: expect.stringContaining("Prefer the message tool for delivery"),
           }),
-        ],
+        ]),
       }),
     );
   });
diff --git a/src/agents/tools/video-generate-background.test.ts b/src/agents/tools/video-generate-background.test.ts
index d03ff6c5d0d..e078aaf92b1 100644
--- a/src/agents/tools/video-generate-background.test.ts
+++ b/src/agents/tools/video-generate-background.test.ts
@@ -148,16 +148,16 @@ describe("video generate background helpers", () => {
           to: "channel:1",
         }),
         expectsCompletionMessage: true,
-        internalEvents: [
+        internalEvents: expect.arrayContaining([
           expect.objectContaining({
             source: "video_generation",
             announceType: "video generation task",
             status: "ok",
             result: expect.stringContaining("MEDIA:/tmp/generated-lobster.mp4"),
             mediaUrls: ["/tmp/generated-lobster.mp4"],
-            replyInstruction: expect.stringContaining("include those exact MEDIA: lines"),
+            replyInstruction: expect.stringContaining("Prefer the message tool for delivery"),
           }),
-        ],
+        ]),
       }),
     );
   });
diff --git a/src/auto-reply/reply/dispatch-from-config.reply-dispatch.test.ts b/src/auto-reply/reply/dispatch-from-config.reply-dispatch.test.ts
index f09ee73c92c..d4e2e55396e 100644
--- a/src/auto-reply/reply/dispatch-from-config.reply-dispatch.test.ts
+++ b/src/auto-reply/reply/dispatch-from-config.reply-dispatch.test.ts
@@ -400,7 +400,7 @@ describe("dispatchReplyFromConfig reply_dispatch hook", () => {
       ctx: createHookCtx(),
       cfg: emptyConfig,
       dispatcher: createDispatcher(),
-      fastAbortResolver: () => ({ handled: false, aborted: false }),
+      fastAbortResolver: async () => ({ handled: false, aborted: false }),
       formatAbortReplyTextResolver: () => "⚙️ Agent was aborted.",
       replyResolver: async () => ({ text: "model reply" }),
     });
@@ -425,7 +425,7 @@ describe("dispatchReplyFromConfig reply_dispatch hook", () => {
       handled: false,
       queuedFinal: false,
       counts: { tool: 0, block: 0, final: 0 },
-    });
+    } satisfies PluginHookReplyDispatchResult);
 
     const result = await dispatchReplyFromConfig({
       ctx: createHookCtx(),