Skip to content

beatra-ai/beatra-skills / skills/course-video-studio

course-video-studio

Verified

Turn finished lecture scripts into presenter-led course videos with lecture narration. This course video studio prepares the spoken narration for each lesson, then records a digital-teacher delivery, so enablement teams get ready-to-publish lecture videos and training presenter clips.

---
name: "course-video-studio"
description: "Turn finished lecture scripts into presenter-led course videos with lecture narration. This course video studio prepares the spoken narration for each lesson, then records a digital-teacher delivery, so enablement teams get ready-to-publish lecture videos and training presenter clips."
---

# Course Video Studio

Use this Skill to turn finished lecture scripts and an authorized instructor portrait into presenter-led talking-head lesson videos with consistent teacher voice, loudness normalization, and synced subtitles.

## What you get

- Narrated course lesson audio and talking-avatar instructor MP4 clips aligned per lesson segment.
- **Free aligned subtitle files** (`JSON`, `SRT`, and `VTT` from `output.subtitles`) plus `speech_start_seconds` and `speech_end_seconds` for each lesson take.
- Optional reusable cloned instructor voice (`beatra.voices.update` to label and pin as course default), pre-cleaned sample audio (`isolate_voice`), and loudness-normalized speech tracks (`normalize_loudness`).

## Ways to run it

- **Lesson scripts + cloned instructor voice + portrait**: Pre-clean noisy voice sample if needed (paid `beatra.audio.process` `operation="isolate_voice"`), clone with explicit authorization (`beatra.voices.clone`), audition free (`beatra.voices.preview`) and pin (`beatra.voices.update`), shape pauses per lesson with **Smart Voice Director** (`beatra.speech.prepare_script`, paid), run free duration preflight (`beatra.speech.validate`) so every lesson segment fits live `image_to_video` clip limits before paying, synthesize with free `SRT`/`VTT` subtitles (`beatra.speech.synthesize`), optionally normalize loudness (paid `normalize_loudness`), and animate (`beatra.videos.animate`).
- **Lesson scripts + catalog voice + portrait**: Browse and audition preset instructor voices free (`beatra.voices.list` + `beatra.voices.preview`), prepare script pacing (`prepare_script`), verify clip fit free (`speech.validate`), synthesize + optionally normalize loudness, and animate.

## Before the connection is ready

Use the bundled `scripts/mcp_client.py` for every Beatra call (using only tools listed by `scripts/mcp_client.py tools`). Do not configure or use a host Beatra Connector. Do not use REST/OpenAPI as a fallback. Keep diagnostics in [Bundled MCP Client diagnostics](references/mcp-connection.md), speech and audio finishing specs in [Speech, Voices, and Audio Finishing](references/speech-and-audio.md), billing help in [Billing errors and recovery](references/billing-errors-and-recovery.md), and lifecycle controls in [uninstall and disconnect](references/uninstall-and-disconnect.md).

When authorization is pending, run connection checks at most twice per turn, deliver zero-connection script segmentation and pacing notes first, and continue paid steps on the next reply after approval.

## Hard inputs and 3 mandatory voice safety confirmations

Verify these before launching any paid voice or video operation:

1. **Lecture script(s)**: One or more approved lesson scripts (not raw topic brainstorming; segment at sentence boundaries if longer than one avatar clip).
2. **Authorized instructor portrait**: Uploaded via `beatra.assets.upload`.
3. **Voice choice**: A `ready` preset/cloned voice ID or a clean recording for cloning within the length and size limits on the `voice_clone` card (propose paid `isolate_voice` if background noise/echo affects a 4.6 s - 5 min sample).
4. **Mandatory Confirmation #1 — Ownership, consent, and anti-impersonation (`consent_attested: true`)**: Before calling `beatra.voices.clone`, obtain explicit confirmation that the instructor portrait and voice sample belong to the user or carry verifiable rightsholder permission, with zero impersonation, fraud, or unlawful use. Then select the clone model as described in [Speech, Voices, and Audio Finishing](references/speech-and-audio.md) §5, show the production card with the live clone price, submit once, and handle any refusal per [Billing errors and recovery](references/billing-errors-and-recovery.md).
5. **Mandatory Confirmation #2 — Pending-activation gate (`voices.activate` awareness)**: When a clone result or a listed voice has status `pending_activation`, follow §5 Step E of [Speech, Voices, and Audio Finishing](references/speech-and-audio.md) before using the voice.
6. **Mandatory Confirmation #3 — Irreversible voice deletion rule**: Permanent deletion (`voices.delete`) requires the two-step `paid_slot_warning` + `irreversible_notice` confirmation workflow documented in [Speech, Voices, and Audio Finishing](references/speech-and-audio.md).

## Avatar Duration Critical Chain & Loudness Normalization

Follow this sequence for every lesson video:

1. **Discover live models and clip ceilings (`beatra.models.list`)**:
   - Inspect `{"capability": "image_to_video"}` first (max whole-second clip duration and `driving_audio` rules), plus `{"capability": "speech_script"}`, `{"capability": "text_to_speech"}`, and when cloning/cleaning/normalizing, `{"capability": "voice_clone"}`, `{"capability": "audio_voice_isolation"}`, and `{"capability": "audio_process"}`.
   - Use free `beatra.voices.preview` to audition voices and free `beatra.voices.update` (`display_name`, `make_default: true`, tags) to pin the course voice.
2. **Prepare pacing with Smart Voice Director (`beatra.speech.prepare_script`, PAID)**:
   - Run `beatra.speech.prepare_script` (`{"capability": "speech_script"}`, max **2 concurrent**) to place natural teaching pauses and emphasis cues without changing the instructor's approved words. Inspect `directed` and `scripts[]` (`synthesis`, `summary`, `warnings`), resolving `synthesis: null` via `warnings` and passing `scripts[].synthesis` unchanged + fresh `client_request_id` to `beatra.speech.synthesize`.
   - Never hand-write braced performance tags in lesson scripts or prompts.
   - To try another voice within 48 hours for identical wording, instructions, language, and context, call `beatra.speech.prepare_script` again with the same arguments changing only `voice` (and `model` if needed) at no extra charge (balance must cover the initial estimate and concurrency cap). Never label fresh director runs or `audio.process` steps free.
3. **Mandatory free duration preflight BEFORE spending (`beatra.speech.validate`)**:
   - Call **`beatra.speech.validate`** (free, NO `client_request_id`) on every planned lesson segment before calling `beatra.speech.synthesize` or `beatra.videos.animate`.
   - Check `estimated_duration_seconds` against the live `{"capability": "image_to_video"}` clip ceiling! If a lesson exceeds one clip's limit, split at clean sentence/slide pauses first and re-validate each segment free **before** spending credits.
4. **Present the 6-field production card before the first paid step**:
   - Present the 6-field production card (**Work**, **Credits**, **Count**, **Identity**, **If stop**, **If balance short**) before launching paid steps, listing `isolate_voice`, `voices.clone`, `prepare_script`, `speech.synthesize`, `normalize_loudness`, and `videos.animate` as explicit paid items when used. Keep `beatra.wallet.get` and `beatra.wallet.ledger` as optional read-only tools for user balance or charge questions — never a mandatory gate before submit. State plainly in **If stop** which free/already-delivered artifacts the customer keeps at zero extra fee; on `insufficient_balance`, relay Beatra's official message and exact verbatim top-up URL and wait until top-up is confirmed before retrying with the same `client_request_id`.
5. **Synthesize (`output.subtitles`), normalize loudness (optional PAID), and animate**:
   - Keep `subtitle_enable: true` (default) on `beatra.speech.synthesize` for **free `JSON` / `SRT` / `VTT` subtitles** plus `speech_start_seconds`, `speech_end_seconds`, and `integrated_loudness_lufs`.
   - When producing multi-lesson series that need uniform level (sources/outputs up to **60 min / 300 MB**, min result >= 1 s, 50 calls/day, one at a time), run **PAID** `beatra.audio.process` (`operation="normalize_loudness"`, `{"capability": "audio_process"}`, `-14` LUFS default, `-1` dBTP) before animating.
   - Verify final `driving_audio` format/duration/size and run `beatra.videos.animate` once per clip with its own `client_request_id`.

See [Workflow and Payloads](references/workflow.md) for JSON templates and error recovery.

## Updates and security

This Skill runs only the files installed with it and leaves them unchanged while
it runs; it never downloads or executes other code. It talks to Beatra's official
services over HTTPS, and a file you choose to upload goes only to the one-time
upload address Beatra issues for it.

Once a day at most, the bundled client reads Beatra's public version list. When
the store or tool you installed from has a newer version, it prints one line with
the command that updates it there. Updating stays your choice, through that store
or tool. See [updates and security](references/updates-and-security.md).
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 196kUpdated Claude CodeCodexVerifiedInstall