Getting Started
Reference
Extend
Help
FAQ
- you can publish your pipes to the screenpipe store and earn money
- set your own pricing model (subscription or one-time purchase)
- publish using the CLI:
bunx --bun @screenpipe/dev@latest pipe publish --name your-pipe \[--paid --price 50\] - our platform handles payment processing and distribution
- get detailed analytics on installations and usage
- check our monetization guide for more details on revenue sharing
- rewind: timeline and search through your screen history
- obsidian: sync your screen activities to obsidian notes
- meeting: automatic meeting notes and summaries
- reddit-auto-posts: create reddit posts based on your screen activities
- linkedin-ai-assistant: helps craft linkedin content
- desktop-to-table: low-level API for desktop automation (similar to playwright but for your desktop)
- visual-conversation: chat with your screen history
- and many more in our pipe store
- create a new pipe with:
bunx --bun @screenpipe/dev@latest pipe create - navigate to your pipe directory:
cd your-pipe-name - start development server:
bun run dev - test your pipe locally in the screenpipe app
- debug and iterate on your implementation
- when ready, build for production:
bun run build - see our pipe development guide for detailed instructions
- minimum requirements:
- dual-core processor (2GHz+)
- 2GB RAM
- 20GB free disk space
- any modern graphics card for hardware acceleration
- recommended specifications:
- quad-core processor or better
- 4GB+ RAM
- 50GB+ free SSD space
- dedicated graphics card with at least 2GB VRAM
- screenpipe is optimized to use hardware acceleration when available, significantly reducing CPU usage
- screenpipe is open source and welcomes contributions
- check our GitHub Issues for tasks to work on
- participate in our bounty program to earn rewards
- join our Discord community to collaborate with other contributors
- follow our contribution guidelines for code standards
- submit pull requests with features, bug fixes, or documentation improvements
- similar to playwright but for your desktop
- provides a low-level API for desktop automation
- 100% accurate interaction with desktop applications
- not based on unreliable pixel-matching like some AI agents
- ideal for building robust automation workflows
- screen recordings are stored as mp4 files
- audio recordings are stored as mp3 files
- all files are saved in $HOME/.screenpipe/ directory
- metadata and indexes are stored in SQLite database
- OCR and transcription data are also indexed in the database
- you can access all data programmatically via our API
- all data is stored in $HOME/.screenpipe/ directory
- to backup, simply copy this directory to another location
- to transfer to another machine, copy the directory to the same location
- you can also use our export API to selectively export data
- database can be accessed directly with SQLite tools
- note: when transferring between different operating systems, some paths may need adjustment
Was this page helpful?