Transcribe And Analyze Agent Skill
Transcribe audio and video from URLs (YouTube, direct media links) using WhisperKit locally. Optionally analyze transcripts with AI when explicitly requested. Use when users provide URLs to media content and request transcription or speech-to-text conversion.
Install
one command, takes just this skill from the repositorynpx skills add https://github.com/nicepkg/ai-workflow --skill transcribe-and-analyzeThe instruction itself
28 sections, as written by the authorTranscribe and Analyze
Local transcription of audio/video content using WhisperKit. Analysis is available on request using OpenAI or local Ollama.
Capabilities
- Transcription - Convert audio/video URLs to text using WhisperKit (runs locally, always available)
- Analysis - Extract insights from transcripts (only when user asks for it, supports OpenAI or Ollama)
Quick Start
Transcribe Only
python3 scripts/transcribe.py "https://youtube.com/watch?v=..."
Transcribe + Analyze (OpenAI)
python3 scripts/transcribe.py "https://youtube.com/watch?v=..."
python3 scripts/analyze_transcript.py whisper-transcriptions/video.md
Transcribe + Analyze (Local)
python3 scripts/transcribe.py "https://youtube.com/watch?v=..."
python3 scripts/analyze_transcript.py whisper-transcriptions/video.md --local
Transcription
Script Options
# Basic
python3 scripts/transcribe.py "URL"
# Custom output directory
python3 scripts/transcribe.py "URL" --output-dir "/path/to/save"
# Higher accuracy (slower)
python3 scripts/transcribe.py "URL" --model medium
# Without timestamps
python3 scripts/transcribe.py "URL" --no-timestamps
# Custom filename
python3 scripts/transcribe.py "URL" --filename "my-transcription.md"
Whisper Models
| Model | Speed | Accuracy | Use Case |
|-------|-------|----------|----------|
| tiny | Fastest | Lowest | Quick drafts, testing |
| base | Fast | Reasonable | Simple content |
| small | Balanced | Good | Default - most use cases |
| medium | Slower | High | Lectures, important content |
| large | Slowest | Highest | Critical accuracy needed |
Dependencies
- yt-dlp -
pip install yt-dlporbrew install yt-dlp - whisperkit-cli - https://github.com/argmaxinc/WhisperKit
Script checks for these and provides install instructions if missing.
Output
Transcriptions save to ./whisper-transcriptions/ as markdown:
# Transcription
**Source:** https://youtube.com/watch?v=example
**Transcribed:** 2025-01-15 14:30:00
**Tool:** WhisperKit
---
[00:00:00.000 --> 00:00:05.000] Welcome to this video...
Analysis
Provider Options
OpenAI API (default):
python3 scripts/analyze_transcript.py transcript.md
python3 scripts/analyze_transcript.py transcript.md --model gpt-4o
Requires OPENAI_API_KEY environment variable.
Local Ollama:
python3 scripts/analyze_transcript.py transcript.md --local
python3 scripts/analyze_transcript.py transcript.md --local --model mistral
Requires Ollama running (ollama serve).
Script Options
# Default comprehensive analysis (OpenAI)
python3 scripts/analyze_transcript.py transcript.md
# Use local Ollama
python3 scripts/analyze_transcript.py transcript.md --local
# Specify model
python3 scripts/analyze_transcript.py transcript.md --model gpt-4o
python3 scripts/analyze_transcript.py transcript.md --local --model llama3.2
# Custom analysis prompt
python3 scripts/analyze_transcript.py transcript.md --prompt "List all tools mentioned"
# Custom output location
python3 scripts/analyze_transcript.py transcript.md --output ~/Documents/analysis.md
# Print to stdout instead of saving
python3 scripts/analyze_transcript.py transcript.md --print
Default Analysis Includes
- Executive summary (2-3 paragraphs)
- Key insights (5-7 bullet points)
- Topics discussed with summaries
- Notable quotes (3-5 memorable quotes)
- Action items and recommendations
- Additional observations
Custom Prompt Examples
--prompt "List all technologies and tools mentioned"
--prompt "What are the main arguments presented?"
--prompt "Extract all statistics and data points"
--prompt "Summarize in 5 bullet points"
--prompt "What questions were asked and how were they answered?"
Dependencies
- openai -
pip install openai(used for both OpenAI and Ollama) - OPENAI_API_KEY environment variable (for OpenAI only)
- Ollama running locally (for --local mode)
Output
Analysis saves alongside transcript as transcript_name_analysis.md:
# Transcript Analysis
**Source Transcript:** path/to/transcript.md
**Analysis Model:** gpt-4o-mini (OpenAI)
**Tokens Used:** 33,763
---
[Analysis content]
Common Workflows
Full Pipeline: URL to Insights (Cloud)
python3 scripts/transcribe.py "https://youtube.com/watch?v=abc123"
python3 scripts/analyze_transcript.py whisper-transcriptions/watch.md
Full Pipeline: URL to Insights (Local)
python3 scripts/transcribe.py "https://youtube.com/watch?v=abc123"
python3 scripts/analyze_transcript.py whisper-transcriptions/watch.md --local
Multiple Analyses on Same Transcript
python3 scripts/analyze_transcript.py transcript.md --output summary.md
python3 scripts/analyze_transcript.py transcript.md --prompt "List action items" --output actions.md
python3 scripts/analyze_transcript.py transcript.md --prompt "Extract quotes" --output quotes.md
Batch Transcription
python3 scripts/transcribe.py "URL1"
python3 scripts/transcribe.py "URL2"
python3 scripts/transcribe.py "URL3"
Reference Files
Troubleshooting (references/troubleshooting.md)
- Download failures
- Transcription errors
- Dependency issues
- API errors
Configuration (references/configuration.md)
- Output format details
- File naming behavior
- Model selection guidance
Usage Patterns (references/usage-patterns.md)
- Common transcription scenarios
- Analysis patterns
- Batch processing tips
Scripts
| Script | Purpose |
|--------|---------|
| scripts/transcribe.py | Download and transcribe audio/video from URLs (WhisperKit) |
| scripts/analyze_transcript.py | AI analysis of transcript files (OpenAI or Ollama) |
Other skills for the same job
different authors, same section of the catalogue
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Improves the quality of images, especially screenshots, by enhancing resolution, sharpness, and clarity. Perfect for preparing images for presentations, documentation, or social media posts.
Create and edit JSON Canvas files (.canvas) with nodes, edges, groups, and connections. Use when working with .canvas files, creating visual canvases, mind maps, flowcharts, or when the user mentions Canvas files in Obsidian.
Publish Markdown articles to X (Twitter) Articles editor with proper formatting. Use when user wants to publish a Markdown file/URL to X Articles, or mentions "publish to X", "post article to Twitter", "X article", or wants help with X Premium article publishing. Handles cover image upload and converts Markdown to rich text automatically.
Downloads videos from YouTube and other platforms for offline viewing, editing, or archival. Handles various formats and quality options.
Teaches the AI to design like a high-end agency. Defines the exact fonts, spacing, shadows, card structures, and animations that make a website feel expensive. Blocks all the common defaults that make AI designs look cheap or generic.
Lightweight WSI tile extraction and preprocessing. Use for basic slide processing tissue detection, tile extraction, stain normalization for H&E images. Best for simple pipelines, dataset preparation, quick tile-based analysis. For advanced spatial proteomics, multiplexed imaging, or deep learning pipelines use pathml.
How to use it
Take nicepkg/transcribe-and-analyze from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.
The instructions reference pip, brew.
Without those the skill loads but fails at the first command.