Skip to main content

Clipcase

Convert screen recording videos of application usage into structured, comprehensive test cases — automatically.

Drop a .mov/.mp4 video, run one command, and get test cases in Markdown, CSV, and Excel formats.


How It Works

Video (.mov/.mp4) ──▶ ffmpeg (frame extraction) ──▶ AI Vision (frame analysis) ──▶ Test Cases (MD/CSV/XLSX)
  1. Frame extraction — Uses ffmpeg to extract screenshots from the video at a configurable frame rate.
  2. AI analysis — Sends sampled frames to a vision-capable LLM (Claude or GPT-4o) to understand the complete user flow.
  3. Test case generation — The AI generates structured test cases following QA best practices, covering positive, negative, and edge case scenarios, along with a self-assessed confidence score.
  4. Export — Outputs test cases in Markdown, CSV, and styled Excel formats.

Prerequisites

Dependency Why Install
Python 3.9+ Runtime python.org
ffmpeg Frame extraction from video brew install ffmpeg (macOS) / sudo apt install ffmpeg (Linux) / choco install ffmpeg (Windows)
API Key AI analysis (Claude or GPT-4o) See API Setup below

Quick Start

1. Clone and install

git clone <repo-url> clipcase
cd clipcase
pip install -r requirements.txt

2. Set your API key

cp .env.example .env
# Edit .env and paste your API key

Or export directly:

export ANTHROPIC_API_KEY=sk-ant-api03-...
# or
export OPENAI_API_KEY=sk-...

3. Run

python -m clipcase your_recording.mov

Prefer a real install instead of running from source? pip install clipcase (or pip install -e . from a clone) registers a clipcase command, so every example below also works as clipcase your_recording.mov — no python -m needed.

That's it. Your test cases will appear as:

  • your_recording_test_cases.md
  • your_recording_test_cases.csv
  • your_recording_test_cases.xlsx

The markdown export and terminal output also include a confidence score — the model's own self-assessment of how completely it captured the recorded flow.


Usage

python -m clipcase <video_file> [options]

Options

Flag Default Description
--fps 2 Frames per second to extract
--provider anthropic LLM provider: anthropic or openai
--output auto Base name for output files
--frames-dir frames Directory for extracted frames
--frames-only off Only extract frames, skip AI
--sample-interval 10 Use every Nth frame for analysis
--principles auto Path to custom principles file

Examples

# Basic — uses Claude, 2 fps
python -m clipcase recording.mov

# Use OpenAI GPT-4o instead
python -m clipcase recording.mov --provider openai

# Use Google Gemini (free tier)
python -m clipcase recording.mov --provider gemini

# Higher frame rate for fast-paced UI
python -m clipcase recording.mov --fps 4

# Just extract frames (no AI cost), review manually
python -m clipcase recording.mov --frames-only

# Custom output name
python -m clipcase recording.mov --output login_flow_tests

# Analyze more frames (lower interval = more frames sent to AI)
python -m clipcase recording.mov --sample-interval 5

API Setup

  1. Go to console.anthropic.com
  2. Create an API key
  3. Add to .env:
    ANTHROPIC_API_KEY=sk-ant-api03-...
    

Option B: OpenAI (GPT-4o)

  1. Go to platform.openai.com
  2. Create an API key
  3. Add to .env:
    OPENAI_API_KEY=sk-...
    
  4. Use --provider openai when running

Option C: Google Gemini — Free tier available

  1. Go to aistudio.google.com/app/apikey
  2. Sign in with your Google account
  3. Click Create API Key (free, no credit card needed)
  4. Add to .env:
    GEMINI_API_KEY=AIza...
    
  5. Use --provider gemini when running

Custom Test Case Principles

You can customize how test cases are written by providing your own principles file:

python -m clipcase recording.mov --principles my_standards.md

If you place a file named test_case_creation_principles.md in the same directory as the video, it will be picked up automatically.

The included test_case_creation_principles.md covers:

  • Clarity, independence, repeatability, traceability, maintainability
  • Naming conventions (Verify that..., Verify if the user...)
  • Pre-conditions, expected outcomes, test data guidelines
  • Test type classification (Smoke, Sanity, Regression, E2E)

Output Formats

Markdown (.md)

Full document with flow analysis, test case table, and summary statistics.

CSV (.csv)

Tab-importable into Google Sheets, Jira, TestRail, or any test management tool. Checkmarks rendered as ✓.

Excel (.xlsx)

Professionally styled with:

  • Color-coded header row
  • Alternating row shading
  • Auto-filters on every column
  • Frozen header row
  • Auto-sized columns

Project Structure

clipcase/
├── clipcase/                           # Main CLI package (core.py, cli.py)
├── requirements.txt                    # Python dependencies
├── .env.example                        # API key template
├── .gitignore                          # Git ignore rules
├── test_case_creation_principles.md    # QA principles (auto-loaded)
├── test_case_template.md               # Reference template
└── README.md                           # This file

Tips

  • Short videos work best: 1–5 minute recordings of a single flow produce the most focused test cases.
  • Record one flow per video: Login flow, checkout flow, onboarding flow — keep them separate.
  • Review and refine: AI-generated test cases are a strong starting point. Have your QA team review and adjust.
  • Lower --sample-interval for complex UIs where many small changes happen quickly.
  • Use --frames-only to preview what the AI will see before spending API credits.

Cost Estimate

Each run makes 2–4 API calls depending on video length:

Provider Approximate Cost per Run
Claude (Anthropic) ~$0.30 – $1.00
GPT-4o (OpenAI) ~$0.50 – $1.50
Gemini 1.5 Pro (Google) Free tier available (generous quota)

Costs scale with frame count and video length.


Troubleshooting

Issue Fix
ffmpeg: command not found Install ffmpeg (see Prerequisites)
ANTHROPIC_API_KEY not set Add key to .env or export in shell
Unicode filename error Tool handles this automatically via symlink
openpyxl not installed Run pip install openpyxl for Excel support
Too few test cases generated Lower --sample-interval to send more frames

License

MIT

Metadata

Release files for clipcase 1.1.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for clipcase 1.1.1
File Size Uploaded Provenance Actions
clipcase-1.1.1.tar.gz 21.7 kB Without provenance Details for clipcase-1.1.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for clipcase 1.1.1
File Interpreter ABI Platform Actions
clipcase-1.1.1-py3-none-any.whl Python 3 none any Details for clipcase-1.1.1-py3-none-any.whl

Total release size: 38.2 kB

Release files / clipcase-1.1.1.tar.gz

Download URL clipcase-1.1.1.tar.gz
Size 21.7 kB
Tags Source
SHA-256 checksum
How to use checksums
ba25a3a7b534b90d40e2c2eacf5c97bb842d59b66c45737b205b55cda8aef525
BLAKE2b-256 checksum
How to use checksums
3c13ed8bbbde5be7df69283c88ccdc67419a907cbf413900e224663c3549ae9d
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.6

Release files / clipcase-1.1.1-py3-none-any.whl

Download URL clipcase-1.1.1-py3-none-any.whl
Size 16.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
d42c73b226c0f83e40a029e769b3eac8e0b288fac590f237b23e1c2ab73ff75e
BLAKE2b-256 checksum
How to use checksums
62a96173c19463b634547f84d7eaa4021563b052fc0a62400ba2abcd6f010ccc
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.6

Security for clipcase 1.1.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page