Gettr Transcribe

Download audio from a GETTR post or streaming page and transcribe it locally with MLX Whisper on Apple Silicon (with timestamps via VTT). Use when given a GE...

Security Scan

The skill's requirements, scripts, and instructions align with its stated purpose (download GETTR media and run a local MLX Whisper transcription); nothing requests unrelated credentials or installs arbitrary remote code.

Purpose & Capability

Name/description (GETTR → local transcription) matches the files and runtime instructions: scripts only download media via ffmpeg and call mlx_whisper to transcribe. Required binaries (ffmpeg, mlx_whisper) are expected for this task and the install metadata lists only pip/brew installs relevant to transcription.

Instruction Scope

SKILL.md instructs obtaining the media URL via browser automation or manual DOM inspection, then running the provided scripts. The instructions do not request unrelated files, system credentials, or broad data collection; they only reference the GETTR URL, derived slug, and local output directory. The skill explicitly excludes gated/authenticated posts.

Install Mechanism

The only install steps are standard package installs: 'brew install ffmpeg' and 'pip install mlx-whisper' (both declared in metadata). There are no downloads from arbitrary/personal URLs or extracted archives in the install spec.

Credentials

No environment variables, credentials, or config paths are required. The only potential external access is to GETTR media URLs (user-supplied) and MLX Whisper may download models from Hugging Face if needed — the README/instructions note possible Hugging Face auth but do not demand credentials.

Persistence & Privilege

The skill does not request permanent presence (always=false), does not modify other skills or system-wide agent settings, and only writes output to a local subdirectory under ./out/gettr-transcribe/<slug> as expected.

Assessment

This skill appears to do what it says: extract a media URL from a GETTR page (you provide or the agent collects via browser automation), download audio with ffmpeg, and transcribe locally with mlx_whisper. Before installing, note: (1) it requires ffmpeg and mlx_whisper on PATH (brew/pip installs are suggested); (2) MLX Whisper may download large models from Hugging Face and could prompt for huggingface-cli login for some models — no credentials are required by default; (3) the agent does not handle private/gated GETTR posts (you must supply a direct media URL for those); (4) transcription is local but downloading media from GETTR means the skill will fetch remote content you point it at; and (5) CPU/ram/disk usage can be significant for large models. If you’re comfortable with those tradeoffs, the skill is internally coherent.

Gettr Transcribe (MLX Whisper)

Quick start

# 1. Parse the slug from the URL (just read it — no script needed)
#    https://gettr.com/post/p1abc2def  → slug = p1abc2def
#    https://gettr.com/streaming/p3xyz → slug = p3xyz

# 2. Get the audio/video URL via browser automation (see Step 1 below)
#    For /streaming/ URLs: extract the .m4a audio URL
#    For /post/ URLs: extract the og:video .m3u8 URL

# 3. Run download + transcription pipeline
bash scripts/run_pipeline.sh "<AUDIO_OR_VIDEO_URL>" "<SLUG>"

To explicitly set the transcription language (recommended for non-English content):

bash scripts/run_pipeline.sh --language zh "<AUDIO_OR_VIDEO_URL>" "<SLUG>"

Common language codes: zh (Chinese), en (English), ja (Japanese), ko (Korean), es (Spanish), fr (French), de (German), ru (Russian).

This outputs:

• ./out/gettr-transcribe/<slug>/audio.wav
• ./out/gettr-transcribe/<slug>/audio.vtt

Summarization is handled separately by the caller (see your prompt for summarization instructions).

---

Workflow (GETTR URL → transcript)

Inputs to confirm

Ask for:

• GETTR post URL
• Language (optional): if the video is non-English and auto-detection fails, ask for the language code (e.g., zh for Chinese)

Notes:

• This skill does not handle authentication-gated GETTR posts.
• This skill does not translate; outputs stay in the video's original language.
• If transcription quality is poor or mixed with English, re-run with explicit --language flag.

Prereqs (local)

• mlx_whisper installed and on PATH
• ffmpeg installed (recommended: brew install ffmpeg)

Step 0 — Parse the slug and pick an output directory

Parse the slug directly from the GETTR URL — just read the last path segment, no script needed:

• https://gettr.com/post/p1abc2def → slug = p1abc2def
• https://gettr.com/streaming/p3xyz789 → slug = p3xyz789

Output directory: ./out/gettr-transcribe/<slug>/

Directory structure:

• ./out/gettr-transcribe/<slug>/audio.wav
• ./out/gettr-transcribe/<slug>/audio.vtt

Step 1 — Get the audio/video URL via browser automation

Use browser automation to navigate to the GETTR URL and extract the media URL from the rendered DOM.

For /streaming/ URLs (primary path)

Streaming pages provide a direct .m4a audio download. Extract it by deriving from the og:video meta tag:

1. Navigate to the GETTR streaming URL and wait for the page to fully load (JavaScript must execute)
2. Extract the audio URL via JavaScript:

   const ogVideo = document.querySelector('meta[property="og:video"]')?.getAttribute("content");
   // Replace .m3u8 with /audio.m4a to get the direct audio download URL
   const audioUrl = ogVideo.replace(".m3u8", "/audio.m4a");
   

3. Use the .m4a URL for the pipeline in Step 2

The .m4a URL is a direct file download (no HLS), so it downloads faster and more reliably than the .m3u8 stream.

For /post/ URLs (fallback path)

Post pages do not have a "Download Audio" button. Extract the og:video URL from the rendered DOM:

1. Navigate to the GETTR post URL and wait for the page to fully load
2. Extract the video URL via JavaScript:

   document.querySelector('meta[property="og:video"]')?.getAttribute("content");
   

3. Use the .m3u8 URL directly for the pipeline in Step 2

If browser automation is not available or fails, see references/troubleshooting.md for how to guide the user to manually extract the URL from their browser.

Step 2 — Run the pipeline (download + transcribe)

Feed the extracted URL and slug into the pipeline:

bash scripts/run_pipeline.sh "<AUDIO_OR_VIDEO_URL>" "<SLUG>"

To explicitly set the language (recommended when auto-detection fails):

bash scripts/run_pipeline.sh --language zh "<AUDIO_OR_VIDEO_URL>" "<SLUG>"

The pipeline does two things:

1. Downloads audio as 16kHz mono WAV via ffmpeg (handles both .m4a and .m3u8 inputs)
2. Transcribes with MLX Whisper, outputting VTT with timestamps

If the pipeline fails with HTTP 412 (stale signed URL)

This error occurs with /streaming/ URLs when the signed URL has expired. Re-run browser automation to get a fresh URL, then retry the pipeline.

If browser automation is not available or fails, see references/troubleshooting.md for how to guide the user to manually extract the fresh URL from their browser.

Notes:

• By default, language is auto-detected. For non-English content where detection fails, use --language.
• If too slow or memory-heavy, try smaller models: mlx-community/whisper-medium or mlx-community/whisper-small.
• If quality is poor, try the full model: mlx-community/whisper-large-v3 (slower but more accurate).
• If --word-timestamps causes issues, the pipeline retries automatically without it.

Bundled scripts

• scripts/run_pipeline.sh: download + transcription pipeline (takes an audio/video URL and slug)
• scripts/download_audio.sh: download/extract audio from HLS (.m3u8) or direct (.m4a) URL to 16kHz mono WAV

Error handling

• No audio track: The download script validates output and reports if the source has no audio.
• HTTP 412 errors: Occurs with /streaming/ URLs when the signed URL has expired. Re-run browser automation to get a fresh URL (see Step 1); if that fails, see references/troubleshooting.md.

Troubleshooting

See references/troubleshooting.md for detailed solutions to common issues including:

• HTTP 412 errors (stale signed URLs)
• Download errors
• Transcription quality issues