[GH-ISSUE #439] Feature Request: Add Screenpipe Integration Tutorial #441

New Issue

GiteaMirror · 2026-04-15T06:53:33-05:00

GiteaMirror commented

2026-04-15 06:53:33 -05:00

Originally created by @louis030195 on GitHub (Jan 25, 2026).
Original GitHub issue: https://github.com/Shubhamsaboo/awesome-llm-apps/issues/439

Feature Request

I'd love to see a tutorial showing how to build LLM apps with Screenpipe context!

What is Screenpipe?

Screenpipe (15k+ ⭐) is an open-source tool that provides 24/7 local screen & mic recording with AI-powered search. It creates a searchable memory of everything you've seen, said, and heard on your computer.

Integration Potential

Screenpipe would be perfect for several tutorial categories:

1. RAG with Screen Context

Build a RAG system that can answer questions about anything the user has seen on their screen:

"What was the name of that person I had a meeting with last Tuesday?"
"Find all the code snippets I looked at today"
"Summarize the articles I read this morning"

2. AI Agent with Personal Context

Create agents that have access to the user's digital history:

Personal assistant that knows your work context
Meeting summarizer that has access to screen content during calls
Research assistant that can recall sources you've viewed

3. Chat with Your Screen History

Similar to "Chat with PDF" but for your entire screen history:

Query weeks of screen recordings with natural language
Extract information from any app you've used
Cross-reference information across different tools

Technical Details

Screenpipe provides:

TypeScript SDK: @screenpipe/js on npm
REST API: Query screen OCR text, audio transcriptions, and more
MCP Server: Model Context Protocol integration for Claude/other LLMs
Local-first: All data stays on user's machine

Example Code

import requests

# Query Screenpipe's local API
response = requests.get("http://localhost:3030/search", params={
    "q": "meeting notes",
    "content_type": "ocr",
    "limit": 10
})
results = response.json()
# Use results as context for your LLM app

Resources

GitHub: https://github.com/mediar-ai/screenpipe
TypeScript SDK: https://www.npmjs.com/package/@screenpipe/js
Documentation: https://docs.screenpi.pe

Would love to contribute a tutorial if there's interest!

Originally created by @louis030195 on GitHub (Jan 25, 2026). Original GitHub issue: https://github.com/Shubhamsaboo/awesome-llm-apps/issues/439 ## Feature Request I'd love to see a tutorial showing how to build LLM apps with **Screenpipe** context! ## What is Screenpipe? [Screenpipe](https://github.com/mediar-ai/screenpipe) (15k+ ⭐) is an open-source tool that provides 24/7 local screen & mic recording with AI-powered search. It creates a searchable memory of everything you've seen, said, and heard on your computer. ## Integration Potential Screenpipe would be perfect for several tutorial categories: ### 1. **RAG with Screen Context** Build a RAG system that can answer questions about anything the user has seen on their screen: - "What was the name of that person I had a meeting with last Tuesday?" - "Find all the code snippets I looked at today" - "Summarize the articles I read this morning" ### 2. **AI Agent with Personal Context** Create agents that have access to the user's digital history: - Personal assistant that knows your work context - Meeting summarizer that has access to screen content during calls - Research assistant that can recall sources you've viewed ### 3. **Chat with Your Screen History** Similar to "Chat with PDF" but for your entire screen history: - Query weeks of screen recordings with natural language - Extract information from any app you've used - Cross-reference information across different tools ## Technical Details Screenpipe provides: - **TypeScript SDK**: `@screenpipe/js` on npm - **REST API**: Query screen OCR text, audio transcriptions, and more - **MCP Server**: Model Context Protocol integration for Claude/other LLMs - **Local-first**: All data stays on user's machine ## Example Code ```python import requests # Query Screenpipe's local API response = requests.get("http://localhost:3030/search", params={ "q": "meeting notes", "content_type": "ocr", "limit": 10 }) results = response.json() # Use results as context for your LLM app ``` ## Resources - GitHub: https://github.com/mediar-ai/screenpipe - TypeScript SDK: https://www.npmjs.com/package/@screenpipe/js - Documentation: https://docs.screenpi.pe Would love to contribute a tutorial if there's interest!

GiteaMirror closed this issue

2026-04-15 06:53:33 -05:00

GiteaMirror commented

2026-04-15 06:53:34 -05:00

@Shubhamsaboo commented on GitHub (Jan 28, 2026):

Sounds interesting. Would love to see the RAG usecase built. Please feel free to create a PR by following the existing contribution guidelines.

@Shubhamsaboo commented on GitHub (Jan 28, 2026): Sounds interesting. Would love to see the RAG usecase built. Please feel free to create a PR by following the existing contribution guidelines.