ASR Stream Rewriter & Translator for Openclaw

An AI-powered skill that aggregates real-time speech-to-text streams and re-writes them into polished, written-form documents or translations upon completion.

hei-maom
v1.0.0
Mar 14, 2026
0
856
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install shumianyu

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install shumianyu using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is ASR Stream Rewriter & Translator?

This skill processes raw Automatic Speech Recognition (ASR) data streams by buffering segments and delaying output until a session is finalized. Instead of producing real-time, fragmented subtitles, it provides a cohesive, grammatically correct written transcript. By leveraging Openclaw Skills, developers can integrate high-quality post-processing for voice inputs, ensuring that oral fillers, repetitions, and self-corrections are removed to create professional-grade documentation.

ASR Stream Rewriter & Translator Use Cases

  • Generating professional meeting minutes from live session transcripts.
  • Cleaning up interview recordings for journalistic or research purposes.
  • Post-processing podcast or video transcripts into readable articles.
  • Refining voice-to-text inputs for technical documentation and emails.

How ASR Stream Rewriter & Translator Works

  1. The skill receives incoming ASR text chunks and ingest metadata.
  2. It appends these chunks into a local buffer while maintaining the overall context.
  3. It monitors for explicit session-end signals such as session_final or end_of_stream markers.
  4. Once the stream ends, it triggers a rewrite engine to convert oral speech into structured written language.
  5. If a target_language is specified, the skill performs a high-fidelity translation of the polished text.
  6. The final, cleaned output is delivered to the user, and the buffer is cleared for the next session.

ASR Stream Rewriter & Translator Setup

To integrate this functionality into your environment, use the standard deployment method for Openclaw Skills:

# Install the skill via the CLI
openclaw install asr-written-rewriter-stream

Configure your environment variables to specify your preferred LLM provider for the rewriting and translation tasks. Ensure your upstream ASR provider is configured to send a finalization flag to trigger the processing logic.

ASR Stream Rewriter & Translator Data Schema & Taxonomy

The skill organizes its processing using a structured input schema to ensure accuracy:

Field Description Requirement
chunk_text The latest segment of text from the ASR engine Required
target_language The language code for optional translation Optional
domain_lexicon A list of specialized terms, names, or jargon Optional
is_session_final Boolean flag indicating the end of the stream Required
recent_context Previous segments used to improve rewriting accuracy Optional

ASR Stream Rewriter & Translator Advanced Features

  • Context-aware rewriting that preserves the speaker's original intent and tone.
  • Domain lexicon integration to ensure technical jargon and proper nouns are identified correctly.
  • Vendor-neutral API abstraction allowing compatibility with any ASR provider.
  • Support for multi-stage processing where translation occurs only after the text is polished.
  • Flexible commit rules based on silence detection or explicit user signals.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*