agent-recipes 0.0.5__py3-none-any.whl

agent_recipes/templates/ai-video-editor/README.md

@@ -0,0 +1,254 @@
+# AI Video Editor Recipe
+
+End-to-end AI-powered video editing that automatically cleans up your videos by removing filler words, repetitions, tangents, and long silences.
+
+**This is a self-contained recipe** - no modifications to praisonai or praisonaiagents required.
+
+## Features
+
+- **Automatic Transcription**: Word-level timestamps using OpenAI Whisper
+- **Filler Word Removal**: Removes "um", "uh", "like", "you know", etc.
+- **Repetition Detection**: Removes stutters and repeated phrases
+- **Tangent Removal**: LLM-based detection of off-topic content
+- **Silence Removal**: Removes long pauses above configurable threshold
+- **Caption Generation**: SRT files or burned-in captions
+- **Edit Decision List**: Professional EDL output for further editing
+
+## Requirements
+
+### System Dependencies
+
+- **FFmpeg**: Required for video processing
+  ```bash
+  # macOS
+  brew install ffmpeg
+  
+  # Linux (Ubuntu/Debian)
+  sudo apt install ffmpeg
+  
+  # Windows
+  # Download from https://ffmpeg.org/download.html
+  ```
+
+### Python Dependencies
+
+```bash
+pip install openai
+```
+
+### Environment Variables
+
+```bash
+export OPENAI_API_KEY="your-key-here"
+```
+
+## Quick Start
+
+### Using Recipe CLI (Recommended)
+
+```bash
+# Navigate to recipe directory
+cd /path/to/Agent-Recipes/agent_recipes/templates/ai-video-editor/src
+
+# Check dependencies
+python -m ai_video_editor.cli doctor
+
+# Edit a video
+python -m ai_video_editor.cli edit input.mp4 --preset podcast --output out.mp4
+
+# Probe video metadata
+python -m ai_video_editor.cli probe input.mp4
+
+# Generate transcript only
+python -m ai_video_editor.cli transcript input.mp4 --output transcript.srt
+```
+
+### Using PraisonAI Templates
+
+```bash
+# List available templates
+praisonai templates list
+
+# Run the template
+praisonai templates run ai-video-editor input.mp4 --output edited.mp4
+```
+
+### Using Python API
+
+```python
+import sys
+sys.path.insert(0, "/path/to/Agent-Recipes/agent_recipes/templates/ai-video-editor/src")
+
+from ai_video_editor import edit, probe, transcript
+
+# Simple edit
+result = edit(
+    input_path="input.mp4",
+    preset="podcast",
+    output_path="out.mp4",
+    verbose=True
+)
+
+print(f"Output: {result.output_path}")
+print(f"Saved: {result.time_saved:.1f}s")
+
+# Custom settings
+result = edit(
+    input_path="meeting.mp4",
+    remove_fillers=True,
+    remove_repetitions=True,
+    remove_tangents=True,
+    target_length="30m",
+    captions="srt",
+    verbose=True
+)
+```
+
+## Presets
+
+| Preset | Fillers | Repetitions | Tangents | Silence Threshold |
+|--------|---------|-------------|----------|-------------------|
+| podcast | ✓ | ✓ | ✗ | 700ms |
+| meeting | ✓ | ✓ | ✓ | 1000ms |
+| course | ✓ | ✓ | ✗ | 500ms |
+| clean | ✓ | ✓ | ✓ | 600ms |
+
+## Output Files
+
+After editing, you'll find these files in the working directory:
+
+- `*_edited.mp4` - Final edited video
+- `transcript.txt` - Plain text transcript
+- `captions.srt` - SRT caption file
+- `edit_plan.json` - Detailed edit plan with segments
+- `edit_decision_list.edl` - Professional EDL file
+- `report.json` - Complete processing report
+
+## CLI Reference
+
+### `edit` Command
+
+```bash
+python -m ai_video_editor.cli edit <input> [options]
+
+Options:
+  --output, -o PATH       Output video path
+  --preset PRESET         Edit preset (podcast, meeting, course, clean)
+  --remove-fillers        Remove filler words
+  --remove-repetitions    Remove repeated phrases
+  --remove-tangents       Remove off-topic content
+  --auto-crop MODE        Crop mode (off, center, face)
+  --target-length TIME    Target duration (e.g., 6m, 90s)
+  --captions MODE         Caption mode (off, srt, burn)
+  --provider PROVIDER     Transcription provider (openai, local, auto)
+  --no-llm                Use simple pattern matching
+  --force                 Overwrite output if exists
+  --json-report PATH      Save JSON report
+  --verbose, -v           Enable verbose output
+```
+
+### `probe` Command
+
+```bash
+python -m ai_video_editor.cli probe <input> [--json]
+```
+
+### `transcript` Command
+
+```bash
+python -m ai_video_editor.cli transcript <input> [options]
+
+Options:
+  --output, -o PATH       Output file path
+  --format FORMAT         Output format (srt, txt, json)
+  --provider PROVIDER     Transcription provider
+  --language LANG         Language code (default: en)
+```
+
+### `doctor` Command
+
+```bash
+python -m ai_video_editor.cli doctor
+```
+
+## How It Works
+
+1. **Probe**: Extract video metadata (duration, resolution, fps)
+2. **Transcribe**: Generate word-level timestamps using Whisper
+3. **Analyze**: LLM identifies segments to remove (or simple pattern matching with --no-llm)
+4. **Optimize**: Snap cuts to word boundaries, apply padding
+5. **Render**: Concatenate kept segments with FFmpeg
+6. **Report**: Generate transcript, captions, and edit report
+
+## Tips
+
+- **Start with podcast preset** for general content
+- **Use meeting preset** for recordings with off-topic discussion
+- **Set target_length** to automatically prioritize cuts
+- **Enable verbose mode** to see progress details
+- **Check edit_plan.json** to review what was removed
+- **Use --no-llm** for faster processing without API calls (after transcription)
+
+## Troubleshooting
+
+### FFmpeg not found
+
+Install FFmpeg for your platform:
+```bash
+# macOS
+brew install ffmpeg
+
+# Ubuntu
+sudo apt install ffmpeg
+```
+
+### API key error
+
+Set your OpenAI API key:
+```bash
+export OPENAI_API_KEY="sk-..."
+```
+
+### Long videos timing out
+
+For videos over 25MB audio, the system automatically chunks the transcription.
+For very long videos (>2 hours), consider splitting first.
+
+### Check all dependencies
+
+```bash
+python -m ai_video_editor.cli doctor
+```
+
+## Architecture
+
+```
+ai-video-editor/
+├── TEMPLATE.yaml          # Template metadata for discovery
+├── README.md              # This file
+├── agents.yaml            # Agent definitions
+├── workflow.yaml          # Workflow definition
+├── requirements.txt       # Python dependencies
+├── src/
+│   └── ai_video_editor/   # Self-contained Python package
+│       ├── __init__.py
+│       ├── __main__.py    # Entry point for python -m
+│       ├── cli.py         # CLI implementation
+│       ├── pipeline.py    # Main orchestrator
+│       ├── config.py      # Presets and configuration
+│       ├── models.py      # Data models
+│       ├── utils.py       # Utilities
+│       ├── ffmpeg_probe.py
+│       ├── transcribe.py
+│       ├── heuristics.py
+│       ├── llm_plan.py
+│       ├── timeline.py
+│       └── render.py
+├── tests/                 # Unit tests
+├── examples/              # Example configurations
+└── scripts/               # Helper scripts
+```
+
+## License
+
+Apache-2.0

agent_recipes/templates/ai-video-editor/TEMPLATE.yaml

@@ -0,0 +1,139 @@
+name: ai-video-editor
+version: "1.0.0"
+description: |
+  End-to-end AI-powered video editor that automatically:
+  - Understands video content through transcription
+  - Removes filler words (um, uh, like, you know)
+  - Removes word repetitions and stutters
+  - Identifies and removes tangent/off-topic segments
+  - Removes long silences
+  - Generates transcripts, captions, and edit decision lists
+  - Produces polished final video output
+  
+  This is a SELF-CONTAINED recipe that runs independently.
+  No modifications to praisonai or praisonaiagents required.
+
+author: praisonai
+license: Apache-2.0
+tags: [video, editing, ai, transcription, filler-removal, podcast, meeting, course]
+
+requires:
+  tools: [shell_tool]
+  packages: [praisonai-tools]
+  env: [OPENAI_API_KEY]
+  tools_sources:
+    - praisonai_tools.video
+  external:
+    - name: ffmpeg
+      check: "ffmpeg -version"
+      install_hint: "brew install ffmpeg (macOS) or apt install ffmpeg (Linux)"
+
+workflow: workflow.yaml
+agents: agents.yaml
+
+config:
+  input:
+    type: string
+    required: true
+    description: Input video file path
+  output:
+    type: string
+    description: Output video file path (default: input_edited.mp4)
+  preset:
+    type: string
+    default: podcast
+    enum: [podcast, meeting, course, clean]
+    description: Edit preset determining removal aggressiveness
+  remove_fillers:
+    type: boolean
+    default: true
+    description: Remove filler words (um, uh, like, etc.)
+  remove_repetitions:
+    type: boolean
+    default: true
+    description: Remove repeated phrases and stutters
+  remove_tangents:
+    type: boolean
+    default: false
+    description: Remove off-topic content (LLM-based)
+  remove_silence:
+    type: boolean
+    default: true
+    description: Remove long silences
+  target_length:
+    type: string
+    description: Target output duration (e.g., 6m, 90s)
+  verbose:
+    type: boolean
+    default: false
+    description: Enable verbose output
+
+defaults:
+  preset: podcast
+  remove_fillers: true
+  remove_repetitions: true
+  remove_tangents: false
+  remove_silence: true
+  verbose: false
+
+cli:
+  command: ai-video-editor
+  args:
+    - name: input
+      positional: true
+      required: true
+      description: Input video file
+    - name: --output
+      short: -o
+      description: Output video file
+    - name: --preset
+      short: -p
+      default: podcast
+      description: Edit preset (podcast, meeting, course, clean)
+    - name: --remove-fillers
+      flag: true
+      description: Remove filler words
+    - name: --remove-repetitions
+      flag: true
+      description: Remove repeated phrases
+    - name: --remove-tangents
+      flag: true
+      description: Remove off-topic content
+    - name: --target-length
+      description: Target duration (e.g., 6m, 90s)
+    - name: --verbose
+      short: -v
+      flag: true
+      description: Enable verbose output
+  examples:
+    - "python -m ai_video_editor.cli edit input.mp4 --preset podcast --output out.mp4"
+    - "python -m ai_video_editor.cli edit input.mp4 --remove-fillers --remove-tangents"
+    - "python -m ai_video_editor.cli probe input.mp4"
+    - "python -m ai_video_editor.cli transcript input.mp4 --output transcript.srt"
+    - "praisonai templates run ai-video-editor input.mp4 --output edited.mp4"
+
+outputs:
+  - name: output_video
+    type: file
+    description: Final edited video file
+  - name: transcript
+    type: file
+    description: Plain text transcript
+  - name: captions
+    type: file
+    description: SRT caption file
+  - name: edit_plan
+    type: file
+    description: JSON edit plan with segments
+  - name: edl
+    type: file
+    description: Edit Decision List file
+  - name: report
+    type: file
+    description: Full JSON report with all metadata
+
+# Self-contained execution
+# This recipe can be run via:
+# 1. praisonai templates run ai-video-editor <input> --output <output>
+# 2. python -m ai_video_editor.cli edit <input> --output <output>
+# 3. ./scripts/run.sh <input> --output <output>

agent_recipes/templates/ai-video-editor/agents.yaml

@@ -0,0 +1,36 @@
+framework: praisonai
+topic: AI Video Editing Pipeline
+roles:
+  video_editor:
+    role: Video Editor
+    goal: Edit videos by removing fillers, repetitions, tangents, and silences
+    backstory: |
+      I am an AI video editor that uses praisonai_tools.video to process videos.
+      I can analyze video content, transcribe audio, identify segments to remove,
+      and render polished final output.
+    tasks:
+      edit_video:
+        description: |
+          Edit the input video using AI analysis:
+          1. Probe video to get metadata
+          2. Transcribe audio with word-level timestamps
+          3. Analyze content to identify segments to remove
+          4. Optimize timeline and render final video
+          
+          Input video: {input}
+          Output path: {output}
+          Preset: {preset}
+          
+          Run the video editing CLI:
+          python -m praisonai_tools.video edit "{input}" --preset {preset} --output "{output}" -v
+        expected_output: |
+          Edited video file with:
+          - Filler words removed
+          - Repetitions removed
+          - Silences trimmed
+          - Transcript and captions generated
+          - Edit decision list created
+    tools:
+      - shell_tool
+
+dependencies: []

agent_recipes/templates/ai-video-editor/requirements.txt

@@ -0,0 +1,8 @@
+# AI Video Editor - Python Dependencies
+# Install with: pip install -r requirements.txt
+
+# Required
+openai>=1.0.0
+
+# Optional - for local transcription
+# faster-whisper>=1.0.0

agent_recipes/templates/ai-video-editor/scripts/run.sh

@@ -0,0 +1,10 @@
+#!/bin/bash
+# AI Video Editor - Shell wrapper
+# Usage: ./run.sh edit input.mp4 --preset podcast --output out.mp4
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+RECIPE_DIR="$(dirname "$SCRIPT_DIR")"
+SRC_DIR="$RECIPE_DIR/src"
+
+# Run the CLI
+cd "$SRC_DIR" && python -m ai_video_editor.cli "$@"

agent_recipes/templates/ai-video-editor/src/ai_video_editor/__init__.py

@@ -0,0 +1,45 @@
+"""
+AI Video Editor - Self-contained recipe for PraisonAI
+
+This module provides end-to-end AI-powered video editing:
+- Transcription with word-level timestamps
+- Filler word detection and removal
+- Repetition detection
+- Tangent/off-topic segment identification
+- Timeline optimization
+- FFmpeg-based rendering
+
+Usage:
+    python -m ai_video_editor.cli edit input.mp4 --preset podcast
+    python -m ai_video_editor.cli probe input.mp4
+    python -m ai_video_editor.cli transcript input.mp4
+"""
+
+__version__ = "1.0.0"
+
+from .pipeline import edit, probe, transcript
+from .config import PRESETS, EditConfig
+from .models import (
+    Word,
+    Segment,
+    SegmentCategory,
+    EditPlan,
+    VideoProbeResult,
+    TranscriptResult,
+    VideoEditResult,
+)
+
+__all__ = [
+    "edit",
+    "probe", 
+    "transcript",
+    "PRESETS",
+    "EditConfig",
+    "Word",
+    "Segment",
+    "SegmentCategory",
+    "EditPlan",
+    "VideoProbeResult",
+    "TranscriptResult",
+    "VideoEditResult",
+]

agent_recipes/templates/ai-video-editor/src/ai_video_editor/__main__.py

@@ -0,0 +1,8 @@
+"""
+Entry point for running as python -m ai_video_editor
+"""
+
+from .cli import main
+
+if __name__ == "__main__":
+    main()