From 97cfa0846c278eb0670e4ff578f1d818bddfb22e Mon Sep 17 00:00:00 2001
From: Peter Steinberger <steipete@gmail.com>
Date: Sat, 17 Jan 2026 03:53:56 +0000
Subject: [PATCH] chore: remove legacy transcription helpers

---
 src/auto-reply/transcription.test.ts | 67 -----------------------
 src/auto-reply/transcription.ts      | 82 ----------------------------
 2 files changed, 149 deletions(-)
 delete mode 100644 src/auto-reply/transcription.test.ts
 delete mode 100644 src/auto-reply/transcription.ts

diff --git a/src/auto-reply/transcription.test.ts b/src/auto-reply/transcription.test.ts
deleted file mode 100644
index bdd97abe4..000000000
--- a/src/auto-reply/transcription.test.ts
+++ /dev/null
@@ -1,67 +0,0 @@
-import fs from "node:fs/promises";
-import os from "node:os";
-import path from "node:path";
-
-import { afterEach, describe, expect, it, vi } from "vitest";
-
-vi.mock("../globals.js", () => ({
-  isVerbose: () => false,
-  logVerbose: vi.fn(),
-  shouldLogVerbose: () => false,
-}));
-
-vi.mock("../process/exec.js", () => ({
-  runExec: vi.fn(),
-}));
-
-const runtime = {
-  error: vi.fn(),
-};
-
-describe("transcribeInboundAudio", () => {
-  afterEach(() => {
-    vi.resetAllMocks();
-    vi.unstubAllGlobals();
-  });
-
-  it("downloads mediaUrl to temp file and returns transcript", async () => {
-    const tmpBuf = Buffer.from("audio-bytes");
-    const tmpFile = path.join(os.tmpdir(), `clawdbot-audio-${Date.now()}.ogg`);
-    await fs.writeFile(tmpFile, tmpBuf);
-
-    const fetchMock = vi.fn(async () => ({
-      ok: true,
-      status: 200,
-      arrayBuffer: async () => tmpBuf,
-    })) as unknown as typeof fetch;
-    vi.stubGlobal("fetch", fetchMock);
-
-    const cfg = {
-      tools: {
-        audio: {
-          transcription: {
-            args: ["echo", "{{MediaPath}}"],
-            timeoutSeconds: 5,
-          },
-        },
-      },
-    };
-    const ctx = { MediaUrl: "https://example.com/audio.ogg" };
-
-    const execModule = await import("../process/exec.js");
-    vi.mocked(execModule.runExec).mockResolvedValue({
-      stdout: "transcribed text\n",
-      stderr: "",
-    });
-    const { transcribeInboundAudio } = await import("./transcription.js");
-    const result = await transcribeInboundAudio(cfg as never, ctx as never, runtime as never);
-    expect(result?.text).toBe("transcribed text");
-    expect(fetchMock).toHaveBeenCalled();
-  });
-
-  it("returns undefined when no transcription command", async () => {
-    const { transcribeInboundAudio } = await import("./transcription.js");
-    const res = await transcribeInboundAudio({ audio: {} } as never, {} as never, runtime as never);
-    expect(res).toBeUndefined();
-  });
-});
diff --git a/src/auto-reply/transcription.ts b/src/auto-reply/transcription.ts
deleted file mode 100644
index e63e79a68..000000000
--- a/src/auto-reply/transcription.ts
+++ /dev/null
@@ -1,82 +0,0 @@
-import crypto from "node:crypto";
-import fs from "node:fs/promises";
-import os from "node:os";
-import path from "node:path";
-
-import type { ClawdbotConfig } from "../config/config.js";
-import { logVerbose, shouldLogVerbose } from "../globals.js";
-import { runExec } from "../process/exec.js";
-import type { RuntimeEnv } from "../runtime.js";
-import { applyTemplate, type MsgContext } from "./templating.js";
-
-const AUDIO_TRANSCRIPTION_BINARY = "whisper";
-
-export function isAudio(mediaType?: string | null) {
-  return Boolean(mediaType?.startsWith("audio"));
-}
-
-export function hasAudioTranscriptionConfig(cfg: ClawdbotConfig): boolean {
-  if (cfg.tools?.audio?.transcription?.args?.length) return true;
-  return Boolean(cfg.audio?.transcription?.command?.length);
-}
-
-export async function transcribeInboundAudio(
-  cfg: ClawdbotConfig,
-  ctx: MsgContext,
-  runtime: RuntimeEnv,
-): Promise<{ text: string } | undefined> {
-  const toolTranscriber = cfg.tools?.audio?.transcription;
-  const legacyTranscriber = cfg.audio?.transcription;
-  const hasToolTranscriber = Boolean(toolTranscriber?.args?.length);
-  if (!hasToolTranscriber && !legacyTranscriber?.command?.length) {
-    return undefined;
-  }
-
-  const timeoutMs = Math.max(
-    (toolTranscriber?.timeoutSeconds ?? legacyTranscriber?.timeoutSeconds ?? 45) * 1000,
-    1_000,
-  );
-  let tmpPath: string | undefined;
-  let mediaPath = ctx.MediaPath;
-  try {
-    if (!mediaPath && ctx.MediaUrl) {
-      const res = await fetch(ctx.MediaUrl);
-      if (!res.ok) throw new Error(`HTTP ${res.status}`);
-      const arrayBuf = await res.arrayBuffer();
-      const buffer = Buffer.from(arrayBuf);
-      tmpPath = path.join(os.tmpdir(), `clawdbot-audio-${crypto.randomUUID()}.ogg`);
-      await fs.writeFile(tmpPath, buffer);
-      mediaPath = tmpPath;
-      if (shouldLogVerbose()) {
-        logVerbose(
-          `Downloaded audio for transcription (${(buffer.length / (1024 * 1024)).toFixed(2)}MB) -> ${tmpPath}`,
-        );
-      }
-    }
-    if (!mediaPath) return undefined;
-
-    const templCtx: MsgContext = { ...ctx, MediaPath: mediaPath };
-    const argv = hasToolTranscriber
-      ? [AUDIO_TRANSCRIPTION_BINARY, ...(toolTranscriber?.args ?? [])].map((part, index) =>
-          index === 0 ? part : applyTemplate(part, templCtx),
-        )
-      : (legacyTranscriber?.command ?? []).map((part) => applyTemplate(part, templCtx));
-    if (shouldLogVerbose()) {
-      logVerbose(`Transcribing audio via command: ${argv.join(" ")}`);
-    }
-    const { stdout } = await runExec(argv[0], argv.slice(1), {
-      timeoutMs,
-      maxBuffer: 5 * 1024 * 1024,
-    });
-    const text = stdout.trim();
-    if (!text) return undefined;
-    return { text };
-  } catch (err) {
-    runtime.error?.(`Audio transcription failed: ${String(err)}`);
-    return undefined;
-  } finally {
-    if (tmpPath) {
-      void fs.unlink(tmpPath).catch(() => {});
-    }
-  }
-}