> [screenpi.pe > computer use AI SDK](https://screenpi.pe/)
# Screenpipe AI SDK
Screenpipe is a computer use AI SDK that securely captures and analyzes computer screen and audio activity. It allows users to install hundreds of AI agents to process this data. Key features include:
* **Secure Screen & Audio Capture:** Records screen and audio activity for analysis.
* **AI Agent Ecosystem:** Supports installation of hundreds of AI agents.
* **Data Visualization:** Provides a data table to visualize recorded data.
* **AI App Store:** Offers a community-built app store with plugins.
* **Time-Based Analysis (Rewind):** Allows users to scroll back in time and ask AI questions about past screen activity.
* **Speaker Identification:** Enables the assignment of names to voices for consistent identification across recordings.
* **Interactive Playground (Example-pipe):** A flexible environment for testing and exploring components, including health status checks, UI recordings, OCR, audio transcription, and screen/audio streaming.
# AI App Store
Discover powerful plugins built by the community.
# Data-Table
Visualize your data in a table, use AI to turn your 24/7 recordings into tables.
# Obsidian
Turn your screen into a living knowledge base with ai analyse your screen/mic activity in real time and write logs, to build a CRM, market research, ideas, user personas, etc. local LLM first.
# Rewind
Scroll back in time, select a time range, cmd+k and ask AI anything about your screen activity or audio conversations.
# Identify-Speakers-v2
Allows you to teach AI to assign names to voices that are then used everywhere else in screenpipe (data, apps, interfaces ...)