fix: add compaction headroom for memory writes

src/auto-reply/reply/agent-runner.heartbeat-typing.test.ts

@@ -164,6 +164,25 @@ describe("runReplyAgent typing (heartbeat)", () => {
     expect(typing.startTypingLoop).not.toHaveBeenCalled();
   });
 
+  it("suppresses partial streaming for NO_REPLY", async () => {
+    const onPartialReply = vi.fn();
+    runEmbeddedPiAgentMock.mockImplementationOnce(
+      async (params: EmbeddedPiAgentParams) => {
+        await params.onPartialReply?.({ text: "NO_REPLY" });
+        return { payloads: [{ text: "NO_REPLY" }], meta: {} };
+      },
+    );
+
+    const { run, typing } = createMinimalRun({
+      opts: { isHeartbeat: false, onPartialReply },
+    });
+    await run();
+
+    expect(onPartialReply).not.toHaveBeenCalled();
+    expect(typing.startTypingOnText).not.toHaveBeenCalled();
+    expect(typing.startTypingLoop).not.toHaveBeenCalled();
+  });
+
   it("starts typing only on deltas in message mode", async () => {
     runEmbeddedPiAgentMock.mockImplementationOnce(async () => ({
       payloads: [{ text: "final" }],

src/auto-reply/reply/agent-runner.ts

@@ -36,7 +36,7 @@ import {
 import { stripHeartbeatToken } from "../heartbeat.js";
 import type { OriginatingChannelType, TemplateContext } from "../templating.js";
 import { normalizeVerboseLevel, type VerboseLevel } from "../thinking.js";
-import { SILENT_REPLY_TOKEN } from "../tokens.js";
+import { isSilentReplyText, SILENT_REPLY_TOKEN } from "../tokens.js";
 import type { GetReplyOptions, ReplyPayload } from "../types.js";
 import {
   createAudioAsVoiceBuffer,
@@ -485,6 +485,7 @@ export async function runReplyAgent(params: {
                       }
                       text = stripped.text;
                     }
+                    if (isSilentReplyText(text, SILENT_REPLY_TOKEN)) return;
                     await typingSignals.signalTextDelta(text);
                     await opts.onPartialReply?.({
                       text,