### Copy .env.example to .env

Source: https://github.com/htekdev/vidpipe/blob/main/docs/configuration.md

Copy the example environment file to start configuring your project.

```bash
cp .env.example .env
```

--------------------------------

### Install Vidpipe from Source

Source: https://github.com/htekdev/vidpipe/blob/main/docs/getting-started.md

Install vidpipe from its source repository. This involves cloning the repo, installing dependencies, building the project, and then starting it.

```bash
git clone https://github.com/htekdev/vidpipe.git
cd vidpipe
npm install
npm run build
npm start
```

--------------------------------

### Install and Run VidPipe

Source: https://github.com/htekdev/vidpipe/blob/main/site/index.md

Install VidPipe globally using npm and run it on a video file to automatically generate edited content, clips, captions, social posts, and a blog draft. This is the quickest way to get started.

```bash
npm install -g vidpipe
vidpipe /path/to/video.mp4
```

--------------------------------

### Example MediaItem Usage

Source: https://github.com/htekdev/vidpipe/blob/main/docs/late-api-reference.md

An example of how to structure the mediaItems array when creating a post.

```json
{
  "mediaItems": [
    { "type": "video", "url": "https://media.getlate.dev/temp/123_abc_video.mp4" }
  ]
}
```

--------------------------------

### Initialize vidpipe Setup

Source: https://github.com/htekdev/vidpipe/blob/main/docs/social-publishing.md

Run the interactive setup wizard for a quick and recommended configuration. This process verifies FFmpeg, sets up API keys, connects social accounts, and creates the schedule.json file.

```bash
vidpipe init
```

--------------------------------

### Platform-Specific Settings Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/late-api-reference.md

An example of how to include platform-specific data, such as title and visibility for YouTube, within the platforms array.

```json
{
  "platforms": [{
    "platform": "youtube",
    "accountId": "acc_123",
    "platformSpecificData": {
      "title": "My Video Title",
      "visibility": "public"
    }
  }]
}
```

--------------------------------

### Minimal VidPipe SDK Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/sdk.md

A basic example demonstrating SDK initialization, running diagnostics, and processing a video. Ensure your OPENAI_API_KEY is set in the environment.

```typescript
import { createVidPipe } from 'vidpipe'

const vidpipe = createVidPipe({
  openaiApiKey: process.env.OPENAI_API_KEY,
})

const diagnostics = await vidpipe.doctor()
if (!diagnostics.allPassed) {
  console.warn(diagnostics.checks)
}

const result = await vidpipe.processVideo('./videos/demo.mp4')
console.log(result.video.videoDir)
console.log(result.shorts.length)
console.log(result.socialPosts.length)
```

--------------------------------

### Set Custom FFmpeg Paths for Windows Manual Install

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Configure FFmpeg and ffprobe binary locations for a manual installation on Windows by setting FFMPEG_PATH and FFPROBE_PATH in a .env file. Ensure the paths point to the 'bin' directory of your FFmpeg installation.

```env
FFMPEG_PATH=C:\tools\ffmpeg\bin\ffmpeg.exe
FFPROBE_PATH=C:\tools\ffmpeg\bin\ffprobe.exe
```

--------------------------------

### Example schedule.json Configuration Snippet

Source: https://github.com/htekdev/vidpipe/blob/main/docs/social-publishing.md

A minimal example of the schedule.json structure, showing timezone and platform-specific posting slots and days to avoid.

```json
{
  "timezone": "America/Chicago",
  "platforms": {
    "linkedin": {
      "slots": [
        { "days": ["tue", "wed"], "time": "08:00", "label": "Morning thought leadership" },
        { "days": ["tue", "wed", "thu"], "time": "12:00", "label": "Lunch break engagement" }
      ],
      "avoidDays": ["sat", "sun"]
    }
  }
}
```

--------------------------------

### Example .env File

Source: https://github.com/htekdev/vidpipe/blob/main/site/guide/configuration.md

Set environment variables for API keys, folder paths, and optional FFmpeg/Exa AI configurations.

```env
OPENAI_API_KEY=sk-your-api-key-here
WATCH_FOLDER=/home/you/Videos/Recordings
OUTPUT_DIR=/home/you/Content/processed

# Optional: explicit FFmpeg paths
# FFMPEG_PATH=/usr/local/bin/ffmpeg
# FFPROBE_PATH=/usr/local/bin/ffprobe

# Optional: Exa AI for web search links in social posts
# EXA_API_KEY=your-exa-api-key-here
```

--------------------------------

### Install FFmpeg via winget (Windows)

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Recommended method for Windows users to install FFmpeg using the Windows Package Manager. Restart your terminal after installation.

```powershell
winget install Gyan.FFmpeg
```

--------------------------------

### VidPipe Configuration Example

Source: https://github.com/htekdev/vidpipe/blob/main/README.md

Example of a .env file for configuring VidPipe. Set environment variables to customize paths, API keys, and other settings.

```env
# .env
OPENAI_API_KEY=sk-your-key-here
WATCH_FOLDER=/path/to/recordings
OUTPUT_DIR=/path/to/output
# EXA_API_KEY=your-exa-key       # Optional: enables web search in social/blog posts
# BRAND_PATH=./brand.json         # Optional: path to brand voice config
# FFMPEG_PATH=/usr/local/bin/ffmpeg
# FFPROBE_PATH=/usr/local/bin/ffprobe
# LATE_API_KEY=sk_your_key_here   # Optional: Late API for social publishing
# GITHUB_TOKEN=ghp_...            # Optional: GitHub token for ID8 idea storage
```

--------------------------------

### FFmpeg Overlay Filter Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research-aspect-ratios.md

Provides an example of the FFmpeg 'overlay' filter for compositing one video stream onto another.

```bash
overlay=main_w-overlay_w:main_h-overlay_h
```

--------------------------------

### Install FFmpeg on Fedora

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Install FFmpeg on Fedora Linux using the dnf package manager.

```bash
sudo dnf install ffmpeg
```

--------------------------------

### Install FFmpeg on Arch Linux

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Install FFmpeg on Arch Linux using the pacman package manager.

```bash
sudo pacman -S ffmpeg
```

--------------------------------

### Full Vidpipe Command Example

Source: https://github.com/htekdev/vidpipe/blob/main/README.md

An example of a comprehensive vidpipe command with multiple options, including watch directory, output directory, API key, brand configuration, and verbose logging.

```bash
vidpipe \
  --watch-dir ~/Videos/Recordings \
  --output-dir ~/Content/processed \
  --openai-key sk-... \
  --brand ./brand.json \
  --verbose
```

--------------------------------

### Check FFmpeg Installation

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Verify that FFmpeg and ffprobe are installed and accessible on your system. Ensure the version is 6.0 or higher.

```bash
ffmpeg -version
ffprobe -version
```

--------------------------------

### Scheduler Configuration Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/scheduling-debug.md

Shows the structure of the schedule.json configuration file, including timezone and platform-specific slot definitions.

```json
{
  "timezone": "America/Chicago",
  "platforms": {
    "tiktok": {
      "slots": [
        { "days": ["tue", "wed", "thu"], "time": "19:00", "label": "Prime entertainment hours" },
        { "days": ["fri", "sat"], "time": "21:00", "label": "Weekend evening" }
      ],
      "avoidDays": []
    }
  }
}
```

--------------------------------

### Debug Log Example Output

Source: https://github.com/htekdev/vidpipe/blob/main/docs/scheduling-debug.md

Example of a debug log message indicating a found available slot for a platform, including the scheduled datetime.

```text
[DEBUG] Found available slot for tiktok: 2026-02-10T19:00:00-06:00
```

--------------------------------

### Constructor-Only Configuration Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/sdk.md

Configure VidPipe solely through options passed to the `createVidPipe` constructor. This is suitable for self-contained applications.

```typescript
import { createVidPipe } from 'vidpipe'

const vidpipe = createVidPipe({
  openaiApiKey: process.env.OPENAI_API_KEY,
  exaApiKey: process.env.EXA_API_KEY,
  youtubeApiKey: process.env.YOUTUBE_API_KEY,
  llmProvider: 'copilot',
  outputDir: './recordings',
  brandPath: './brand.json',
})
```

--------------------------------

### Preact Setup and Platform Configuration

Source: https://github.com/htekdev/vidpipe/blob/main/src/L7-app/review/public/index.html

Initializes Preact, htm for JSX-like syntax, and defines platform-specific configurations.

```javascript
import { h, render } from 'https://esm.sh/preact@10'
import { useState, useEffect, useCallback, useRef } from 'https://esm.sh/preact@10/hooks'
import htm from 'https://esm.sh/htm@3'
const html = htm.bind(h)

// ── Platform config ──
const PLATFORMS = {
  tiktok: { icon: '🎵', color: '#00f2ea', name: 'TikTok' },
  youtube: { icon: '▶️', color: '#ff0000', name: 'YouTube' },
  instagram: { icon: '📸', color: '#e1306c', name: 'Instagram' },
  linkedin: { icon: '💼', color: '#0077b5', name: 'LinkedIn' },
  twitter: { icon: '🐦', color: '#1da1f2', name: 'X/Twitter' },
}
```

--------------------------------

### Mock Simplification Example: With Layers

Source: https://github.com/htekdev/vidpipe/blob/main/docs/architecture/layers.md

Demonstrates how the layered architecture reduces the number of necessary mocks. By adhering to layer boundaries, only direct dependencies (L2 clients in this L3 service example) need to be mocked.

```typescript
vi.mock('../../L2-clients/whisper/whisperClient.js')
vi.mock('../../L2-clients/ffmpeg/audioExtraction.js')
```

--------------------------------

### Basic .env Configuration

Source: https://github.com/htekdev/vidpipe/blob/main/site/guide/configuration.md

For most users, a simple `.env` file with the OpenAI API key is sufficient. Other settings default to sensible values. Use `vidpipe init` for interactive setup.

```env
OPENAI_API_KEY=sk-your-key-here
```

--------------------------------

### Verify vidpipe Setup

Source: https://github.com/htekdev/vidpipe/blob/main/docs/social-publishing.md

Run the doctor command to check if your vidpipe installation and configuration are set up correctly.

```bash
vidpipe doctor
```

--------------------------------

### Full Vidpipe Configuration Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/getting-started.md

Run vidpipe with a comprehensive set of options, including watch directory, output directory, API keys, branding, and verbose logging.

```bash
vidpipe \
  --watch-dir ~/Videos/Recordings \
  --output-dir ~/Content/processed \
  --openai-key sk-... \
  --exa-key exa-... \
  --brand ./my-brand.json \
  --verbose
```

--------------------------------

### ASS Subtitle Formatting Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research-captions.md

Demonstrates how to format ASS dialogue lines to align with the start of the first word and the end of the last word, with comments indicating silence periods.

```ass
; Group 1: words from 0.24s to 1.58s (then silence until 2.10s)
Dialogue: 0,0:00:00.24,0:00:01.58,Hormozi,,0,0,0,,In the tapestry of time
; Group 2: words from 2.10s to 4.97s
Dialogue: 0,0:00:02.10,0:00:04.97,Hormozi,,0,0,0,,a visitor from shadows past
; (screen is blank from 1.58s to 2.10s)
```

--------------------------------

### Example .env file configuration

Source: https://github.com/htekdev/vidpipe/blob/main/docs/configuration.md

A sample .env file showing how to set environment variables for vidpipe, including API keys and directory paths. Uncomment lines to configure FFmpeg paths explicitly.

```env
OPENAI_API_KEY=sk-your-api-key-here
WATCH_FOLDER=/home/you/Videos/Recordings
OUTPUT_DIR=/home/you/Content/processed

# Optional: explicit FFmpeg paths
# FFMPEG_PATH=/usr/local/bin/ffmpeg
```

--------------------------------

### Install FFmpeg via Homebrew (macOS)

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Install FFmpeg on macOS using the Homebrew package manager. This command installs both ffmpeg and ffprobe and adds them to your system's PATH.

```bash
brew install ffmpeg
```

--------------------------------

### Complete Post Example (Multi-Platform with Media)

Source: https://github.com/htekdev/vidpipe/blob/main/docs/late-api-reference.md

This JSON structure demonstrates how to create a post with content and media items, targeting multiple social media platforms with platform-specific configurations. It includes scheduling information.

```json
{
  "content": "Excited to announce our new product! 🎉",
  "mediaItems": [
    { "url": "https://media.getlate.dev/temp/123_product.mp4", "type": "video" }
  ],
  "platforms": [
    {
      "platform": "youtube",
      "accountId": "yt_123",
      "platformSpecificData": {
        "title": "New Product Announcement",
        "visibility": "public",
        "categoryId": "28",
        "firstComment": "Thanks for watching! 🔔"
      }
    },
    {
      "platform": "tiktok",
      "accountId": "tt_456",
      "platformSpecificData": {
        "tiktokSettings": {
          "privacy_level": "PUBLIC_TO_EVERYONE",
          "allow_comment": true,
          "allow_duet": true,
          "allow_stitch": true,
          "content_preview_confirmed": true,
          "express_consent_given": true
        }
      }
    },
    {
      "platform": "instagram",
      "accountId": "ig_789",
      "platformSpecificData": {
        "firstComment": "Link in bio! 🔗"
      }
    },
    {
      "platform": "linkedin",
      "accountId": "li_012",
      "platformSpecificData": {
        "firstComment": "What do you think? 👇"
      }
    },
    {
      "platform": "twitter",
      "accountId": "tw_345"
    }
  ],
  "scheduledFor": "2024-11-01T10:00:00Z",
  "timezone": "America/New_York"
}
```

--------------------------------

### Install FFmpeg via MacPorts (macOS)

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Install FFmpeg on macOS using the MacPorts package manager.

```bash
sudo port install ffmpeg
```

--------------------------------

### Example Queue Structure

Source: https://github.com/htekdev/vidpipe/blob/main/site/guide/social-publishing.md

Illustrates the directory structure for posts in the publishing queue, including video, metadata, and content files.

```tree
publish-queue/
  ├── my-tip-tiktok/
  │   ├── media.mp4        # Platform-optimized video
  │   ├── metadata.json    # Scheduling and platform data
  │   └── post.md          # Post text content

```

--------------------------------

### Install FFmpeg on Ubuntu/Debian

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Install FFmpeg on Ubuntu or Debian-based Linux distributions using the apt package manager.

```bash
sudo apt update
sudo apt install ffmpeg
```

--------------------------------

### SDK Method: startInterview

Source: https://github.com/htekdev/vidpipe/blob/main/specs/L7-app/commands/ideateStart.md

Initiates an interview session programmatically via the SDK. It takes an idea number and options, and provides real-time event callbacks.

```APIDOC
## startInterview(ideaNumber, options)

### Description
Starts an interactive interview session for a given idea number using the SDK. It allows for real-time event observation and input handling.

### Parameters
#### Parameters
- **ideaNumber** (string) - Required - The GitHub Issue number of the idea to start an interview for.
- **options** (object) - Optional - Configuration options for the interview session.
  - **mode** (string) - Optional - The mode of the interview. Defaults to `interview`.
  - **progress** (boolean) - Optional - If true, enables structured JSONL interview events to stderr.
  - **onEvent** (function) - Optional - Callback function to receive real-time interview events.
  - **answerProvider** (function) - Optional - Callback function called when the agent requires user input.
```

--------------------------------

### Install FFmpeg via Chocolatey (Windows)

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Alternative installation method for Windows using the Chocolatey package manager.

```powershell
choco install ffmpeg
```

--------------------------------

### Environment-Driven Configuration Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/sdk.md

Configure VidPipe using environment variables. The SDK is initialized without arguments, relying on the environment for settings.

```bash
OPENAI_API_KEY=sk-...
EXA_API_KEY=exa_...
YOUTUBE_API_KEY=AIza...
OUTPUT_DIR=./recordings
BRAND_PATH=./brand.json
LLM_PROVIDER=copilot
```

```typescript
import { createVidPipe } from 'vidpipe'

const vidpipe = createVidPipe()
```

--------------------------------

### Install Vidpipe Globally

Source: https://github.com/htekdev/vidpipe/blob/main/docs/getting-started.md

Install the vidpipe CLI globally using npm. This makes the `vidpipe` command available system-wide.

```bash
npm install -g vidpipe
```

--------------------------------

### Install Dependencies for ONNX Runtime and Image Processing

Source: https://github.com/htekdev/vidpipe/blob/main/docs/screen-object-detection.md

Install the necessary npm packages for running ONNX models and processing images.

```bash
npm install onnxruntime-node sharp
```

--------------------------------

### FFmpeg CropDetect Filter Example

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research-aspect-ratios.md

Example of using FFmpeg's 'cropdetect' filter to automatically detect black borders in a video.

```bash
cropdetect=24:16:0
```

--------------------------------

### CLI Command: ideate start

Source: https://github.com/htekdev/vidpipe/blob/main/specs/L7-app/commands/ideateStart.md

Initiates an interactive interview session to develop a content idea. Accepts a GitHub Issue number as a required argument and supports an optional --mode flag.

```APIDOC
## ideate start <issue-number>

### Description
Prepares a content idea for recording by running an interactive interview session in the terminal. Supports Socratic questioning to help develop the idea through iterative Q&A.

### Method
CLI Command

### Parameters
#### Path Parameters
- **issue-number** (string) - Required - The GitHub Issue number associated with the idea.

#### Options
- **--mode** (string) - Optional - Selects the session mode. Defaults to `interview`.
- **--progress** (boolean) - Optional - Enables structured JSONL interview events to stderr.
```

--------------------------------

### CI/CD or Automation Configuration

Source: https://github.com/htekdev/vidpipe/blob/main/docs/configuration.md

Ideal for automated workflows, this setup processes a single video, outputs to a temporary directory, and requires an OpenAI API key to be set as an environment variable. The --once flag ensures it runs and exits.

```bash
OPENAI_API_KEY=sk-... vidpipe \
  --once \
  --output-dir /tmp/output \
  /path/to/video.mp4
```

--------------------------------

### Install VidPipe as a Library (SDK)

Source: https://context7.com/htekdev/vidpipe/llms.txt

Install the vidpipe package as a dependency for programmatic use in your Node.js project. Ensure your project uses ESM.

```bash
npm install vidpipe
```

--------------------------------

### Transcription and summary only

Source: https://github.com/htekdev/vidpipe/blob/main/docs/configuration.md

This example demonstrates how to process a video for transcription and summary only, by skipping all optional pipeline stages using CLI flags.

```bash
vidpipe \
  --no-silence-removal \
  --no-shorts \
  --no-social \
  --no-captions \
  /path/to/video.mp4
```

--------------------------------

### Install Gemini Node.js SDK

Source: https://github.com/htekdev/vidpipe/blob/main/docs/gemini-video-understanding.md

Install the Google GenAI SDK for Node.js using npm. This is a prerequisite for using the Gemini API in your project.

```bash
npm install @google/genai
```

--------------------------------

### Example Multi-Line Layout

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research-captions.md

An example of how multi-line captions are laid out on a vertical screen. Both lines appear simultaneously, with only the currently spoken word highlighted.

```plaintext
Example multi-line layout (vertical 1080×1920):
Line 1: "In the tapestry"      (centered at y=1400)
Line 2: "of time"              (centered at y=1480)
```

--------------------------------

### Initialize and Process Video with Vidpipe SDK

Source: https://github.com/htekdev/vidpipe/blob/main/README.md

Import and initialize the Vidpipe SDK, then use it to process a video file. Ensure your OpenAI API key is set as an environment variable. The output directory for recordings can be specified during initialization.

```typescript
import { createVidPipe } from 'vidpipe'

const vidpipe = createVidPipe({
  openaiApiKey: process.env.OPENAI_API_KEY,
  outputDir: './recordings',
})

const result = await vidpipe.processVideo('./videos/demo.mp4', {
  skipGit: true,
})

console.log(result.video.videoDir)
console.log(result.shorts.length)
```

--------------------------------

### System Prompt Guidance: Layouts

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research/producer-agent.md

Provides guidance for the ProducerAgent's system prompt regarding layout selection, including defaults and specific use cases.

```markdown
### Layout Guidance
- Start with `split_layout` as default
- `only_webcam` for personal moments (intros, reactions, emotional beats)
- `only_screen` for detailed technical content (code walkthroughs, demos)
- `zoom_screen` always requires prior `capture_frame` + region specification
```

--------------------------------

### Recommended Pipeline for Smart Cropping

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research-aspect-ratios.md

This pseudocode outlines the recommended steps for a smart cropping pipeline, including webcam detection, region extraction, and stacking.

```pseudocode
1. Detect webcam overlay region (OpenCV face detection in corners)
2. If webcam detected:
   a. Extract webcam region → scale for bottom panel
   b. Remove webcam region from screen content
   c. Smart-crop screen content for top panel
   d. Stack vertically (vstack)
3. If no webcam:
   a. Center-crop the most active region
   b. Scale to 9:16
4. Apply captions overlay (for accessibility + engagement)
```

--------------------------------

### Example JSONL Progress Events

Source: https://github.com/htekdev/vidpipe/blob/main/README.md

These are examples of the self-contained JSON objects emitted as progress events. Each line represents a distinct event in the pipeline.

```jsonl
{"event":"pipeline:start","videoPath":"video.mp4","totalStages":16,"timestamp":"..."}
{"event":"stage:start","stage":"ingestion","stageNumber":1,"totalStages":16,"name":"Ingestion","timestamp":"..."}
{"event":"stage:complete","stage":"ingestion","stageNumber":1,"totalStages":16,"name":"Ingestion","duration":423,"success":true,"timestamp":"..."}
{"event":"stage:skip","stage":"shorts","stageNumber":7,"totalStages":16,"name":"Shorts","reason":"SKIP_SHORTS","timestamp":"..."}
{"event":"pipeline:complete","totalDuration":45000,"stagesCompleted":14,"stagesFailed":0,"stagesSkipped":2,"timestamp":"..."}
```

--------------------------------

### Configure Environment Variables

Source: https://github.com/htekdev/vidpipe/blob/main/README.md

Copy the example environment file and edit it to add your required API keys. The OpenAI API key is mandatory for transcription and AI features.

```bash
# Unix/Mac
cp .env.example .env
# Windows (PowerShell)
Copy-Item .env.example .env

# Then edit .env and add your OpenAI API key (REQUIRED):
#   OPENAI_API_KEY=sk-your-key-here
```

--------------------------------

### Hoisted Mock Setup with vi.hoisted()

Source: https://github.com/htekdev/vidpipe/blob/main/audit/mockability/README.md

The `vi.hoisted()` pattern is consistently used across test files for ESM-compatible mock setup, especially for complex dependencies.

```typescript
vi.hoisted()
```

--------------------------------

### Create Split-Screen Video (Screen + Webcam)

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research-aspect-ratios.md

Combines screen content and webcam feed into a 9:16 portrait video. Screen content occupies the top 65% and webcam the bottom 35%. Requires FFmpeg with filter_complex support.

```bash
ffmpeg -i input.mp4 -filter_complex "
  [0:v]crop=ih*16/9:ih:0:0,scale=1080:1248[screen];
  [0:v]crop=480:270:iw-490:ih-280,scale=1080:672[webcam];
  [screen][webcam]vstack=inputs=2[out]"
-map "[out]" -map 0:a -c:v libx264 -preset medium -crf 23 -c:a aac output_split.mp4
```

--------------------------------

### Vidpipe Progress Event Examples

Source: https://context7.com/htekdev/vidpipe/llms.txt

Examples of structured JSONL events emitted to stderr for pipeline progress. These events include pipeline start/complete, stage start/complete/skip/error.

```jsonl
{"event":"pipeline:start","videoPath":"video.mp4","totalStages":16,"timestamp":"2026-03-01T10:00:00Z"}
```

```jsonl
{"event":"stage:start","stage":"ingestion","stageNumber":1,"totalStages":16,"name":"Ingestion","timestamp":"..."}
```

```jsonl
{"event":"stage:complete","stage":"ingestion","stageNumber":1,"totalStages":16,"name":"Ingestion","duration":423,"success":true,"timestamp":"..."}
```

```jsonl
{"event":"stage:skip","stage":"shorts","stageNumber":6,"totalStages":16,"name":"Shorts","reason":"SKIP_SHORTS","timestamp":"..."}
```

```jsonl
{"event":"stage:error","stage":"blog","stageNumber":13,"totalStages":16,"name":"Blog","error":"Network timeout","timestamp":"..."}
```

```jsonl
{"event":"pipeline:complete","totalDuration":45231,"stagesCompleted":14,"stagesFailed":1,"stagesSkipped":1,"timestamp":"..."}
```

--------------------------------

### FFmpeg Burn-In Command Examples

Source: https://github.com/htekdev/vidpipe/blob/main/docs/research-captions.md

Demonstrates FFmpeg commands for burning ASS captions into video. Includes basic usage, specifying a custom fonts directory, and high-quality encoding settings.

```bash
# Basic burn-in
ffmpeg -i input.mp4 -vf "ass=captions.ass" -c:a copy output.mp4

# With custom fonts directory
ffmpeg -i input.mp4 -vf "ass=captions.ass:fontsdir=./fonts/" -c:a copy output.mp4

# High quality encoding
ffmpeg -i input.mp4 \
  -vf "ass=captions.ass:fontsdir=./fonts/" \
  -c:v libx264 -preset slow -crf 18 \
  -c:a copy \
  output.mp4
```

--------------------------------

### Install FFmpeg in Alpine Linux (Docker)

Source: https://github.com/htekdev/vidpipe/blob/main/docs/ffmpeg-setup.md

Install FFmpeg within an Alpine Linux environment, typically used in Docker containers, using the apk package manager.

```dockerfile
RUN apk add --no-cache ffmpeg
```

--------------------------------

### Initialize VidPipe SDK

Source: https://github.com/htekdev/vidpipe/blob/main/docs/sdk.md

Create an instance of the VidPipe SDK. Configuration options can be passed to the constructor.

```typescript
const vidpipe = createVidPipe(config?)
```

--------------------------------

### Run Vidpipe Directly with npx

Source: https://github.com/htekdev/vidpipe/blob/main/docs/getting-started.md

Execute vidpipe directly using npx without a global installation. Useful for quick tests or when you don't want to install globally.

```bash
npx vidpipe --once /path/to/video.mp4
```