microsoft/autogen · error · ValueError

audio_output_path must end with .mp3.

Error message

audio_output_path must end with .mp3.

What it means

extract_audio in video_surfer hard-codes MP3 output (ffmpeg format=mp3) and enforces it by rejecting any audio_output_path that does not end in .mp3 (case-insensitive). Allowing arbitrary extensions would let callers write other file types, so the extension acts as both a format contract and a write-scope guard (paired with a cwd path-traversal check).

Source

Thrown at python/packages/autogen-ext/src/autogen_ext/agents/video_surfer/tools.py:32

def extract_audio(video_path: str, audio_output_path: str) -> str:
    """
    Extracts audio from a video file and saves it as an MP3 file.

    :param video_path: Path to the video file (must be a local file path, not a URL).
    :param audio_output_path: Path to save the extracted audio file (must end with .mp3).
    :return: Confirmation message with the path to the saved audio file.
    """
    import os
    import re

    # Reject URLs to prevent SSRF via ffmpeg
    if re.match(r"^[a-zA-Z][a-zA-Z0-9+\-.]*://", video_path):
        raise ValueError("video_path must be a local file path, not a URL.")

    # Enforce .mp3 extension to prevent writing arbitrary file types
    if not audio_output_path.lower().endswith(".mp3"):
        raise ValueError("audio_output_path must end with .mp3.")

    # Prevent path traversal — output must stay within the current working directory
    cwd = os.path.realpath(os.getcwd())
    output_real = os.path.realpath(audio_output_path)
    if not output_real.startswith(cwd + os.sep) and output_real != cwd:
        raise ValueError("audio_output_path must be within the current working directory.")

    (ffmpeg.input(video_path).output(audio_output_path, format="mp3").run(quiet=True, overwrite_output=True))  # type: ignore
    return f"Audio extracted and saved to {audio_output_path}."


def transcribe_audio_with_timestamps(audio_path: str) -> str:
    """
    Transcribes the audio file with timestamps using the Whisper model.

    :param audio_path: Path to the audio file.
    :return: Transcription with timestamps.
    """

View on GitHub (pinned to 027ecf0a37)

Solutions

1. Rename the output to end with .mp3: pass "/tmp/out/clip.mp3".
2. If you need another format, call ffmpeg directly (ffmpeg.input(...).output(path, format=...)) rather than this tool.
3. In agent prompts, specify the output path convention *.mp3 for this tool.

Example fix

# before
extract_audio("/tmp/work/clip.mp4", "/tmp/work/clip.wav")

# after
extract_audio("/tmp/work/clip.mp4", "/tmp/work/clip.mp3")

import os

def mp3_output(path: str) -> str:
    root, _ = os.path.splitext(path)
    return root + ".mp3"

audio_out = mp3_output(audio_output_path)  # always .mp3

def is_mp3_path(p: str) -> bool:
    return isinstance(p, str) and p.lower().endswith(".mp3")

When it happens

Trigger: extract_audio("clip.mp4", "/tmp/out/clip.wav") — any output path whose lowercase form does not end with .mp3, including .MP3-safe variants like clip.mp4 or clip.

Common situations: LLM tool calls proposing .wav/.m4a outputs; pipelines that derive the output name from the input name without rewriting the extension.

Related errors

• audio_output_path must be within the current working directo
• video_path must be a local file path, not a URL.
• Timestamp {timestamp:.2f}s is out of range [0s, {duration:.2
• Failed to list MCP prompts
• Failed to get MCP prompt

AI-assisted analysis of microsoft/autogen@027ecf0a37 (2026-08-15). Data as JSON: /api/errors/3a2b98b12f7cb727. Report an issue: GitHub.