Introduction

In the DeepSeek Harness (DSH) development workflow, adding speech recognition capabilities to an agent usually requires manually integrating an API. A ready-made plugin that can directly handle multilingual speech-to-text conversion can reduce redundant work. xby-asr-1 is exactly such a plugin.

Plugin Overview

Name: xby-asr-1
Maintainer: xby-skill
Category: Model Inference
Positioning: General-purpose speech recognition, supporting multiple major languages as well as minor languages.

Core Features

The plugin provides the following features:

  1. set_xby_apikey
    Sets an API key in the chat. This action is automatically persisted and remains valid after a restart.
  2. asr_whisper_small
    General-purpose speech recognition, supporting multiple major languages as well as minor languages. Requires an audio file URL as input.
  3. asr_whisper_small_for_data_base64
    General-purpose speech recognition, supporting multiple major languages as well as minor languages. Requires the Base64-encoded content of an audio file as input.
  4. asr_whisper_small_for_data_file
    General-purpose speech recognition, supporting multiple major languages as well as minor languages. Requires a local file path to an audio file as input.

Installation and Enablement

Use the following command to install the plugin:

dsh plugin --profile web add github:xby_skill/xby-asr-1

Typical Usage

  1. Obtain the key: Register on the XiaoBenYang official website and obtain an API key.
  2. Configure the key: In the DSH chat environment, invoke the set_xby_apikey command and enter the key you obtained.
  3. Perform recognition: Choose the appropriate feature based on how the audio is provided:
    - If it is a network URL, use asr_whisper_small.
    - If it is Base64-encoded data, use asr_whisper_small_for_data_base64.
    - If it is a local file path, use asr_whisper_small_for_data_file.

Use Cases and Notes

This plugin is suitable for agent developers who need to integrate speech recognition capabilities. Note: the plugin runs with the permissions of the current dsh process. Before installation, please review the source code and the license (MIT).

Conclusion

xby-asr-1 provides basic ASR capabilities. For more configuration details, refer to the community directory or the GitHub repository.

Community Directory
GitHub Repository