arisa 3.1.8 → 3.2.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "arisa",
-  "version": "3.1.8",
+  "version": "3.2.0",
   "description": "Telegram + Pi Agent modular assistant",
   "type": "module",
   "main": "src/index.js",

package/src/core/artifacts/normalize-for-reasoning.js

@@ -1,7 +1,9 @@
 function mimeMatches(pattern, mimeType = "") {
-  if (!pattern || !mimeType) return false;
-  if (pattern === mimeType) return true;
-  if (pattern.endsWith("/*")) return mimeType.startsWith(`${pattern.slice(0, -2)}/`);
+  const normalizedMimeType = mimeType.split(";")[0].trim().toLowerCase();
+  const normalizedPattern = pattern?.trim().toLowerCase();
+  if (!normalizedPattern || !normalizedMimeType) return false;
+  if (normalizedPattern === normalizedMimeType) return true;
+  if (normalizedPattern.endsWith("/*")) return normalizedMimeType.startsWith(`${normalizedPattern.slice(0, -2)}/`);
   return false;
 }
 
@@ -20,8 +22,9 @@ function looksLikeAudioTranscriptionTool(tool) {
 }
 
 export function shouldNormalizeArtifactToText(artifact, desiredMimeType = "text/plain") {
+  const mimeType = artifact?.mimeType?.split(";")[0].trim().toLowerCase();
   return desiredMimeType === "text/plain"
-    && (artifact?.mimeType?.startsWith("audio/") || artifact?.mimeType?.startsWith("video/"));
+    && (mimeType?.startsWith("audio/") || mimeType?.startsWith("video/"));
 }
 
 export function selectPipeTool({ toolRegistry, artifact, desiredMimeType }) {

package/src/transport/telegram/media.js

@@ -6,6 +6,28 @@ async function downloadToBuffer(ctx, fileId) {
   return Buffer.from(await response.arrayBuffer());
 }
 
+function mimeTypeFromAudioFileName(fileName = "") {
+  const extension = fileName.toLowerCase().split(".").pop();
+  return {
+    flac: "audio/flac",
+    m4a: "audio/mp4",
+    mp3: "audio/mpeg",
+    mp4: "audio/mp4",
+    mpeg: "audio/mpeg",
+    mpga: "audio/mpga",
+    ogg: "audio/ogg",
+    opus: "audio/ogg",
+    wav: "audio/wav",
+    webm: "audio/webm"
+  }[extension] || "";
+}
+
+function normalizeDocumentMimeType(document) {
+  const mimeType = document.mime_type || "";
+  if (mimeType && mimeType !== "application/octet-stream") return mimeType;
+  return mimeTypeFromAudioFileName(document.file_name) || mimeType || "application/octet-stream";
+}
+
 function incomingCaptionMetadata(ctx) {
   return ctx.message?.caption ? { caption: ctx.message.caption } : {};
 }
@@ -33,6 +55,26 @@ export async function captureIncomingArtifact(ctx, artifactStore) {
     });
   }
 
+  if (ctx.message?.audio) {
+    const audio = ctx.message.audio;
+    const fileName = audio.file_name || `${chatId}-${ctx.msg.message_id}`;
+    const content = await downloadToBuffer(ctx, audio.file_id);
+    return store.createGeneratedFile({
+      fileName,
+      content,
+      kind: "audio",
+      mimeType: audio.mime_type || "audio/mpeg",
+      source: baseSource,
+      metadata: {
+        duration: audio.duration,
+        performer: audio.performer,
+        title: audio.title,
+        fileSize: audio.file_size,
+        ...incomingCaptionMetadata(ctx)
+      }
+    });
+  }
+
   if (ctx.message?.video) {
     const video = ctx.message.video;
     const fileName = video.file_name || `${chatId}-${ctx.msg.message_id}.mp4`;
@@ -54,13 +96,15 @@ export async function captureIncomingArtifact(ctx, artifactStore) {
   }
 
   if (ctx.message?.document) {
-    const fileName = ctx.message.document.file_name || `${chatId}-${ctx.msg.message_id}`;
-    const content = await downloadToBuffer(ctx, ctx.message.document.file_id);
+    const document = ctx.message.document;
+    const fileName = document.file_name || `${chatId}-${ctx.msg.message_id}`;
+    const mimeType = normalizeDocumentMimeType(document);
+    const content = await downloadToBuffer(ctx, document.file_id);
     return store.createGeneratedFile({
       fileName,
       content,
-      kind: "document",
-      mimeType: ctx.message.document.mime_type || "application/octet-stream",
+      kind: mimeType.startsWith("audio/") ? "audio" : "document",
+      mimeType,
       source: baseSource,
       metadata: incomingCaptionMetadata(ctx)
     });

package/tools/openai-transcribe/index.js

@@ -8,8 +8,55 @@ import { getToolConfigPath } from "../../src/runtime/paths.js";
 const toolName = "openai-transcribe";
 const config = await loadToolConfig(toolName, defaults);
 
+const supportedUploadExtensions = new Set([
+  ".flac",
+  ".m4a",
+  ".mp3",
+  ".mp4",
+  ".mpeg",
+  ".mpga",
+  ".ogg",
+  ".wav",
+  ".webm"
+]);
+
+const mimeUploadExtensions = new Map([
+  ["audio/aac", ".m4a"],
+  ["audio/flac", ".flac"],
+  ["audio/m4a", ".m4a"],
+  ["audio/mp3", ".mp3"],
+  ["audio/mp4", ".m4a"],
+  ["audio/mpeg", ".mp3"],
+  ["audio/mpga", ".mpga"],
+  ["audio/ogg", ".ogg"],
+  ["audio/opus", ".ogg"],
+  ["audio/wav", ".wav"],
+  ["audio/wave", ".wav"],
+  ["audio/webm", ".webm"],
+  ["audio/x-m4a", ".m4a"],
+  ["audio/x-wav", ".wav"],
+  ["video/mp4", ".mp4"],
+  ["video/webm", ".webm"]
+]);
+
+function baseMimeType(mimeType = "") {
+  return mimeType.split(";")[0].trim().toLowerCase();
+}
+
+function uploadFileNameForArtifact(artifact) {
+  const currentName = path.basename(artifact.path);
+  const currentExtension = path.extname(currentName).toLowerCase();
+  if (supportedUploadExtensions.has(currentExtension)) return currentName;
+
+  const extension = mimeUploadExtensions.get(baseMimeType(artifact.mimeType));
+  if (!extension) return currentName;
+
+  const parsed = path.parse(currentName);
+  return `${parsed.name || "audio"}${extension}`;
+}
+
 function printHelp() {
-  console.log(`openai-transcribe\n\nUsage:\n  node index.js --help\n  node index.js run --request-file <json>\n\nExpected input:\n  {\n    "artifact": { "path": "/abs/media.ogg", "mimeType": "audio/ogg" },\n    "args": {}\n  }\n\nConfig at ${getToolConfigPath(toolName)}:\n  OPENAI_API_KEY\n  MODEL\n`);
+  console.log(`openai-transcribe\n\nUsage:\n  node index.js --help\n  node index.js run --request-file <json>\n\nExpected input:\n  {\n    "artifact": { "path": "/abs/media.ogg", "mimeType": "audio/ogg" },\n    "args": {}\n  }\n\nSupported upload formats include flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, and webm.\n\nConfig at ${getToolConfigPath(toolName)}:\n  OPENAI_API_KEY\n  MODEL\n`);
 }
 
 async function run(requestFile) {
@@ -32,7 +79,7 @@ async function run(requestFile) {
   await stat(artifact.path);
   const form = new FormData();
   const data = await readFile(artifact.path);
-  form.append("file", new Blob([data]), path.basename(artifact.path));
+  form.append("file", new Blob([data], { type: baseMimeType(artifact.mimeType) || "application/octet-stream" }), uploadFileNameForArtifact(artifact));
   form.append("model", config.MODEL);
 
   const response = await fetch("https://api.openai.com/v1/audio/transcriptions", {

package/tools/openai-transcribe/tool.manifest.json

@@ -2,7 +2,24 @@
   "name": "openai-transcribe",
   "description": "Transcribe audio files and video audio tracks with OpenAI audio transcription API.",
   "entry": "index.js",
-  "input": ["audio/ogg", "audio/mpeg", "audio/wav", "audio/mp4", "video/mp4"],
+  "input": [
+    "audio/aac",
+    "audio/flac",
+    "audio/m4a",
+    "audio/mp3",
+    "audio/mp4",
+    "audio/mpeg",
+    "audio/mpga",
+    "audio/ogg",
+    "audio/opus",
+    "audio/wav",
+    "audio/wave",
+    "audio/webm",
+    "audio/x-m4a",
+    "audio/x-wav",
+    "video/mp4",
+    "video/webm"
+  ],
   "output": ["text/plain"],
   "configSchema": {
     "OPENAI_API_KEY": {