GitHub Copilot CLI — agentic coding with screen context

Connect screenpipe to GitHub Copilot CLI via MCP so Copilot can reference your screen history, meeting transcriptions, and app context from the terminal.

GitHub Copilot CLI is GitHub’s terminal-based AI coding agent. it supports MCP servers, so you can connect screenpipe and let Copilot reference what you’ve been seeing, hearing, and doing.

setup

before setting up, make sure screenpipe is running on your machine:

npx screenpipe@latest record

check that it’s running:

curl http://localhost:3030/health

the easiest way to connect is the built-in /mcp add slash command inside Copilot CLI:

/mcp add

fill in the prompts:

field	value
name	screenpipe
type	local
command	npx
args	-y screenpipe-mcp
tools	*

or edit ~/.copilot/mcp-config.json directly:

{
  "mcpServers": {
    "screenpipe": {
      "type": "local",
      "command": "npx",
      "args": ["-y", "screenpipe-mcp"],
      "tools": ["*"]
    }
  }
}

restart Copilot CLI after editing the config.

stdio vs http — which should I use?

use stdio (type: "local" in config). this is the standard for local MCP servers on the same machine. screenpipe runs locally on your machine, so stdio communication has no network overhead and is the intended pattern.

HTTP transport is only for remote scenarios where your MCP client (Claude, Cursor, etc.) runs on a different machine than screenpipe — see openclaw for that pattern. for Copilot CLI on the same machine, always choose stdio.

verify connection

inside Copilot CLI, run:

/mcp

you should see screenpipe listed with its tools.

available tools

tool	description
search-content	search accessibility-first screen text, audio transcriptions, and input events with time/app filters
export-video	create video exports from screen recordings
activity-summary	compressed activity overview for a time range
list-meetings	list detected meetings with duration, app, and attendees
search-elements	search structured UI elements (accessibility tree nodes and OCR text blocks)
frame-context	get accessibility text, parsed tree nodes, and extracted URLs for a specific frame

usage examples

ask Copilot CLI to use screenpipe naturally:

> what was I working on this morning?

> find the error message I saw in my terminal earlier

> summarize the kubernetes docs I was reading

> export a video of my screen from 2-3pm today

> what did the team discuss in standup about the API refactor?

> which apps did I switch between most today?

example workflows

recall context from earlier:

> I was reading a blog post about rust async earlier today,
> search screenpipe and summarize the key points

reference meeting discussion:

> search my audio transcriptions for what was discussed in standup
> about the API refactor, then help me implement it

debug with screen history:

> I saw an error message flash on screen, search screenpipe
> to find it and help me fix the issue

requirements

• screenpipe running on localhost:3030 (start with npx screenpipe@latest record)
• GitHub Copilot CLI installed
• Node.js >= 18.0.0

troubleshooting

how do I check I’m on the latest screenpipe CLI?

always invoke screenpipe with the @latest npm tag — it pulls the freshest build from npm each time:

npx screenpipe@latest record

check the version you’re running:

npx screenpipe@latest --version

MCP not connecting?

1. verify screenpipe is running:

   curl http://localhost:3030/health
   

   you should see {"status":"healthy"}.

2. confirm Copilot sees the server:

   /mcp
   

3. remove and re-add via /mcp if the server is in an error state.

queries returning empty?

• check screenpipe has data: curl "http://localhost:3030/search?limit=1"
• ensure screen recording permissions are granted
• give screenpipe a minute or two of running before querying

permission errors on macOS?

• System Settings → Privacy & Security → Screen Recording → enable for your terminal

still stuck? ask in our discord.