screenpipe
Local screen and audio recording with natural language search, feeding your activity history to an agent via MCP
High risk
We rate an entry high when the tool writes to external systems, handles money, production databases or secrets, or runs arbitrary commands. The CLI installs it only with your consent.
Why this level
- Continuously records screen, audio and keystrokes, including potentially sensitive data
- Needs careful filter setup to avoid capturing passwords and personal data
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/screenpipeDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
This entry is high risk, so there is no one-click install. Review the code and add the config by hand.
Run in a terminal
claude mcp add --transport stdio screenpipe -- npx -y screenpipe-mcp@latestOr add to the file .mcp.json, in the project
{
"mcpServers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcpServers key.
Add to the file ~/.cursor/mcp.json, for all projects
{
"mcpServers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcpServers key. For a single project, put the same block into .cursor/mcp.json.
Run in a terminal
code --add-mcp '{"name":"screenpipe","type":"stdio","command":"npx","args":["-y","screenpipe-mcp@latest"]}'Or add to the file .vscode/mcp.json, in the project
{
"servers": {
"screenpipe": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the servers key.
Run in a terminal
codex mcp add screenpipe -- npx -y screenpipe-mcp@latestOr add to the file ~/.codex/config.toml, for all projects
[mcp_servers.screenpipe]
command = "npx"
args = ["-y", "screenpipe-mcp@latest"]If the file already exists, append the block to the end.
Add to the file ~/.gemini/settings.json, for all projects
{
"mcpServers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcpServers key.
Add to the file ~/.config/devin/mcp_config.json, for all projects
{
"mcpServers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcpServers key. Legacy Cascade keeps the MCP config in ~/.codeium/windsurf/mcp_config.json.
Formerly Windsurf.
Add to the file cline_mcp_settings.json, for all projects
{
"mcpServers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcpServers key. Open the settings file in Cline: MCP Servers tab, Configure MCP Servers.
Add to the file .roo/mcp.json, in the project
{
"mcpServers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcpServers key.
A fork of Roo Code, same .roo folders.
Add to the file opencode.json, in the project
{
"mcp": {
"screenpipe": {
"type": "local",
"command": [
"npx",
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcp key.
Add to the file ~/.config/zed/settings.json, for all projects
{
"context_servers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the context_servers key.
Add to the file .codeassistant/mcp.json, in the project
{
"mcpServers": {
"screenpipe": {
"command": "npx",
"args": [
"-y",
"screenpipe-mcp@latest"
]
}
}
}If the file already exists, add the server inside the mcpServers key.
Install and start recording: npx screenpipe record. Connect MCP: claude mcp add screenpipe -- npx -y screenpipe-mcp@latest, or run npx screenpipe setup to install the skills and MCP into every detected agent.
Other ways from the author
npx screenpipe recordStarts local recording via the CLI, without the desktop app.
This is third-party code. Review the repository files before installing.
What it does
screenpipe continuously records your screen and audio locally: it captures the accessibility tree, runs OCR as a fallback, transcribes speech by speaker, and logs keystrokes and app switches. Everything stays on your machine, with filters for windows, apps and passwords, and optional encryption at rest. The screenpipe-mcp server gives an agent access to that history, so you can ask Claude what you saw in the last 5 minutes or have it build an automation based on repeated actions.
Who it is for. For anyone who wants an agent to remember the context of their computer work and find any moment without manual screenshotting.
Good fit when
- You need to quickly find what was on screen or said in a call an hour ago
- You want an agent to build automations based on what you actually do at the computer
- You need a local alternative to cloud activity trackers with no data sent out
Not a fit when
- You are not comfortable with continuous screen and audio recording to disk, even locally
- You are short on disk space: storage runs around 20gb per month
Example request
What did I see on screen in the last 5 minutes and what was said in the call?Limitations
Uses noticeable resources: 5-10 percent CPU and 0.5-3gb RAM continuously in the background, plus around 20gb of storage per month. The desktop app is source-available rather than a strict OSI open source license.
How to disable. Stop recording and remove the app or package: npx screenpipe stop, then remove screenpipe-mcp from your MCP config and uninstall the desktop app.
MCP
- Transport
- stdio
- Authentication
- not required
Security check
- Continuously records screen, audio and keystrokes, including potentially sensitive data
- Needs careful filter setup to avoid capturing passwords and personal data
README in short
The README presents screenpipe as local screen and audio capture that feeds agents context about what you saw, heard and did. Install via the auto-updating desktop app, or via npx screenpipe record followed by npx screenpipe setup, which installs MCP and skills into every detected agent. It also documents a CLI-only mode for coding agents with the screenpipe-cli skill. Specs: accessibility tree capture, OCR fallback, speaker-labeled transcription, 5-10 percent CPU, 20gb of storage per month, optional encryption.
FAQ
Does data go to the cloud?
No, everything is stored locally on your machine and it works offline.
Can I exclude passwords and specific apps from recording?
Yes, there are filters for windows, apps, Chrome extensions, and a model that recognizes passwords and personal data.
Related
A self-hosted knowledge base with block-level references and a built-in MCP server for connecting AI agents to your notes
A CLI for every Google Workspace API with JSON output and agent skills: Drive, Gmail, Calendar, Sheets and more
Local search over Markdown notes, docs and meeting transcripts: keywords, semantic search and reranking, with an MCP server
A task manager for AI-driven development: breaks a PRD into dependent tasks and guides the agent through them via MCP or CLI