# Transcribe Audio with Azure AI and Python

Speech transcription projects require correct Azure client setup and workflow choices. This skill provides Python examples for batch and real-time transcription, authentication, diarization, and language selection.

## Install

```bash
npx skillstore add sickn33/azure-ai-transcription-py
```

## Metadata

- Status: approved
- Slug: sickn33-azure-ai-transcription-py
- Skillstore revision: r2
- Version status: missing
- Tree hash: b50f7c93a12cf50081b2bf35a2873a475318d08d8ceef42f44880873eec10fd4
- Author: sickn33
- GitHub username: sickn33
- License: MIT
- Repository: https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/azure-ai-transcription-py
- Ref: f9e2c34b4f19c7f3e6b0a1e93227b5f77cc12526
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network, env\_access
- Quality score: 79
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/sickn33-azure-ai-transcription-py
- Manifest: https://skillstore.pages.dev/api/skills/sickn33-azure-ai-transcription-py/manifest

## Capabilities

- Shows package installation and environment variable setup for an Azure endpoint and subscription key.
- Constructs a Python transcription client with key-based authentication.
- Starts batch transcription from a hosted audio URL with locale and speaker diarization settings.
- Starts a streaming session, sends a local audio file, and iterates recognized text events.
- Provides workflow guidance for language selection, timestamps, backpressure, and session cleanup.

## Use Cases

- Transcribe Support Calls: Create a Python batch workflow for stored call recordings, selected language, and speaker separation.
- Process Research Interviews: Prepare hosted interview audio for batch transcription with diarization and timestamp planning.
- Stream Meeting Audio: Build a real-time transcription flow that sends local audio and processes text events.

## Prompt Templates

### Set Up the Client

```
Show me how to install Azure AI Transcription for Python and configure key authentication with environment variables.
```

### Create a Batch Job

```
Create a batch transcription workflow for [audio URL] using locale [locale]. Enable diarization and explain each configuration choice.
```

### Build a Streaming Flow

```
Design a real-time Python transcription flow for [audio file]. Include event processing, backpressure handling, session cleanup, and error handling.
```

### Plan a Production Pipeline

```
Design a production transcription pipeline for [workload]. Compare batch and streaming, then address locales, diarization, timestamps, retries, secrets, and result storage.
```

## Limitations

- Requires an Azure transcription resource, subscription key, network access, and the Python client package.
- Does not support DefaultAzureCredential according to the skill documentation.
- Examples omit production error handling, retries, result persistence, and detailed response processing.
- Does not verify service availability, package versions, account permissions, or transcription accuracy.

## Best Practices

- Keep the subscription key in an environment variable and never include it in prompts, logs, or source control.
- Specify the expected language and enable diarization only when the recording contains multiple speakers.
- Use batch jobs for long hosted files, and manage backpressure and cleanup for streaming sessions.

## Anti Patterns

- Do not embed subscription keys directly in Python files or share them with transcription output.
- Do not use streaming for long stored recordings when a batch job is more suitable.
- Do not assume default language, unlimited stream capacity, or automatic session cleanup.

## Security Audit

- Audited at: 2026-07-23T19:50:27.915\+00:00
- Summary: The nine external-command alerts match Markdown code fences, not Ruby or shell backtick execution. The URLs are explicit Azure and storage placeholders, while environment access is limited to documented client configuration. No prompt injection, covert exfiltration, or malicious intent was found.

## Stats

- Views: 105
- Downloads: 13
- Favorites: 0
- Popularity score: 0