LinkedIn Agent MCP

Through this LinkedIn Agent MCP, AI assistants like Claude can connect to your LinkedIn. Access profiles and companies, search for jobs and people, manage saved jobs, update job-search profile settings, and inspect analytics.

Attribution

This project was originally bootstrapped from Daniel Sticker's LinkedIn MCP work and has since been extended into a broader LinkedIn automation and job-search manager. Credit for the original foundation goes to Daniel Sticker and the original linkedin-mcp-server project.

Related MCP server: LinkedIn Custom MCP Server

Installation Methods

👁 uvx
👁 Docker
👁 Install DXT Extension
👁 Development

https://github.com/user-attachments/assets/eb84419a-6eaf-47bd-ac52-37bc59c83680

Usage Examples

Research the background of this candidate https://www.linkedin.com/in/ayushkumar-exl/

Get this company profile for partnership discussions https://www.linkedin.com/company/inframs/

Suggest improvements for my CV to target this job posting https://www.linkedin.com/jobs/view/4252026496

What has Anthropic been posting about recently? https://www.linkedin.com/company/anthropicresearch/

Features & Tool Status

Structured Helper Fields

Some tools return additive structured fields alongside the existing raw text output.

• search_jobs keeps sections.search_results and also returns jobs, a structured list with title, company, location, job_id, and url when those values can be resolved from the current LinkedIn DOM.

• search_people and get_company_people return paginated results arrays with normalized PersonCard fields and filters_applied / warnings metadata. search_people.match_mode controls whether the tool stays strict, broadens automatically, or runs a broad company/background search immediately.

• get_saved_jobs and get_job_recommendations return paginated jobs arrays with normalized JobCard fields.

• Profile-write tools (update_profile_headline, set_open_to_work, add_profile_skills, set_featured_skills) support preview-first flows via dry_run or confirm=false and return structured write envelopes with additive data.

• get_my_post_analytics returns the standard read envelope and exposes parsed posts at data.posts. Each post object includes author, url, text_preview, time_ago, reactions, comments, reposts, and impressions.

🚀 uvx Setup (Recommended - Universal)

Prerequisites: Install uv and run uvx patchright install chromium to set up the browser.

Installation

Step 1: Create a session (first time only)

uvx linkedin-scraper-mcp --login

This opens a browser for you to log in manually (5 minute timeout for 2FA, captcha, etc.). The browser profile is saved to ~/.linkedin-mcp/profile/.

Step 2: Client Configuration:

{
 "mcpServers": {
 "linkedin": {
 "command": "uvx",
 "args": ["linkedin-scraper-mcp"]
 }
 }
}

uvx Setup Help

Transport Modes:

• Default (stdio): Standard communication for local MCP servers

• Streamable HTTP: For web-based MCP server

• If no transport is specified, the server defaults to stdio

• An interactive terminal without explicit transport shows a chooser prompt

CLI Options:

• --login - Open browser to log in and save persistent profile

• --no-headless - Show browser window (useful for debugging scraping issues)

• --log-level {DEBUG,INFO,WARNING,ERROR} - Set logging level (default: WARNING)

• --transport {stdio,streamable-http} - Optional: force transport mode (default: stdio)

• --host HOST - HTTP server host (default: 127.0.0.1)

• --port PORT - HTTP server port (default: 8000)

• --path PATH - HTTP server path (default: /mcp)

• --logout - Clear stored LinkedIn browser profile

• --timeout MS - Browser timeout for page operations in milliseconds (default: 5000)

• --user-data-dir PATH - Path to persistent browser profile directory (default: ~/.linkedin-mcp/profile)

• --chrome-path PATH - Path to Chrome/Chromium executable (for custom browser installations)

Basic Usage Examples:

# Create a session interactively
uvx linkedin-scraper-mcp --login

# Run with debug logging
uvx linkedin-scraper-mcp --log-level DEBUG

HTTP Mode Example (for web-based MCP clients):

uvx linkedin-scraper-mcp --transport streamable-http --host 127.0.0.1 --port 8080 --path /mcp

Runtime server logs are emitted by FastMCP/Uvicorn.

Test with mcp inspector:

1. Install and run mcp inspector bunx @modelcontextprotocol/inspector

2. Click pre-filled token url to open the inspector in your browser

3. Select Streamable HTTP as Transport Type

4. Set URL to http://localhost:8080/mcp

5. Connect

6. Test tools

Installation issues:

• Ensure you have uv installed: curl -LsSf https://astral.sh/uv/install.sh | sh

• Check uv version: uv --version (should be 0.4.0 or higher)

Session issues:

• Browser profile is stored at ~/.linkedin-mcp/profile/

• Make sure you have only one active LinkedIn session at a time

Login issues:

• LinkedIn may require a login confirmation in the LinkedIn mobile app for --login

• You might get a captcha challenge if you logged in frequently. Run uvx linkedin-scraper-mcp --login which opens a browser where you can solve it manually.

Timeout issues:

• If pages fail to load or elements aren't found, try increasing the timeout: --timeout 10000

• Users on slow connections may need higher values (e.g., 15000-30000ms)

• Can also set via environment variable: TIMEOUT=10000

Custom Chrome path:

• If Chrome is installed in a non-standard location, use --chrome-path /path/to/chrome

• Can also set via environment variable: CHROME_PATH=/path/to/chrome

🐳 Docker Setup

Prerequisites: Make sure you have Docker installed and running.

Authentication

Docker runs headless (no browser window), so you need to create a browser profile locally first and mount it into the container.

Step 1: Create profile using uvx (one-time setup)

uvx linkedin-scraper-mcp --login

This opens a browser window where you log in manually (5 minute timeout for 2FA, captcha, etc.). The browser profile is saved to ~/.linkedin-mcp/profile/.

Step 2: Configure Claude Desktop with Docker

{
 "mcpServers": {
 "linkedin": {
 "command": "docker",
 "args": [
 "run", "--rm", "-i",
 "-v", "~/.linkedin-mcp:/home/pwuser/.linkedin-mcp",
 "iushv/linkedin-agent-mcp:latest"
 ]
 }
 }
}

Docker Setup Help

Transport Modes:

• Default (stdio): Standard communication for local MCP servers

• Streamable HTTP: For a web-based MCP server

• If no transport is specified, the server defaults to stdio

• An interactive terminal without explicit transport shows a chooser prompt

CLI Options:

• --log-level {DEBUG,INFO,WARNING,ERROR} - Set logging level (default: WARNING)

• --transport {stdio,streamable-http} - Optional: force transport mode (default: stdio)

• --host HOST - HTTP server host (default: 127.0.0.1)

• --port PORT - HTTP server port (default: 8000)

• --path PATH - HTTP server path (default: /mcp)

• --logout - Clear stored LinkedIn browser profile

• --timeout MS - Browser timeout for page operations in milliseconds (default: 5000)

• --user-data-dir PATH - Path to persistent browser profile directory (default: ~/.linkedin-mcp/profile)

• --chrome-path PATH - Path to Chrome/Chromium executable (rarely needed in Docker)

HTTP Mode Example (for web-based MCP clients):

docker run -it --rm \
 -v ~/.linkedin-mcp:/home/pwuser/.linkedin-mcp \
 -p 8080:8080 \
 iushv/linkedin-agent-mcp:latest \
 --transport streamable-http --host 0.0.0.0 --port 8080 --path /mcp

Runtime server logs are emitted by FastMCP/Uvicorn.

Test with mcp inspector:

1. Install and run mcp inspector bunx @modelcontextprotocol/inspector

2. Click pre-filled token url to open the inspector in your browser

3. Select Streamable HTTP as Transport Type

4. Set URL to http://localhost:8080/mcp

5. Connect

6. Test tools

Docker issues:

• Make sure Docker is installed

• Check if Docker is running: docker ps

Login issues:

• Make sure you have only one active LinkedIn session at a time

• LinkedIn may require a login confirmation in the LinkedIn mobile app for --login

• You might get a captcha challenge if you logged in frequently. Run uvx linkedin-scraper-mcp --login which opens a browser where you can solve captchas manually. See the uvx setup for prerequisites.

Timeout issues:

• If pages fail to load or elements aren't found, try increasing the timeout: --timeout 10000

• Users on slow connections may need higher values (e.g., 15000-30000ms)

• Can also set via environment variable: TIMEOUT=10000

Custom Chrome path:

• If Chrome is installed in a non-standard location, use --chrome-path /path/to/chrome

• Can also set via environment variable: CHROME_PATH=/path/to/chrome

📦 Claude Desktop (DXT Extension)

Prerequisites: Claude Desktop and Docker installed & running

One-click installation for Claude Desktop users:

1. Download the DXT extension

2. Double-click to install into Claude Desktop

3. Create a session: uvx linkedin-scraper-mcp --login

DXT Extension Setup Help

First-time setup timeout:

• Claude Desktop has a ~60 second connection timeout

• If the Docker image isn't cached, the pull may exceed this timeout

• Fix: Pre-pull the image before first use:

  docker pull iushv/linkedin-agent-mcp:2.3.0

• Then restart Claude Desktop

Docker issues:

• Make sure Docker is installed

• Check if Docker is running: docker ps

Login issues:

• Make sure you have only one active LinkedIn session at a time

• LinkedIn may require a login confirmation in the LinkedIn mobile app for --login

• You might get a captcha challenge if you logged in frequently. Run uvx linkedin-scraper-mcp --login which opens a browser where you can solve captchas manually. See the uvx setup for prerequisites.

Timeout issues:

• If pages fail to load or elements aren't found, try increasing the timeout: --timeout 10000

• Users on slow connections may need higher values (e.g., 15000-30000ms)

• Can also set via environment variable: TIMEOUT=10000

🐍 Local Setup (Develop & Contribute)

Prerequisites: Git and uv installed

Installation

# 1. Clone repository
git clone https://github.com/iushv/linkedin-agent-mcp
cd linkedin-agent-mcp

# 2. Install UV package manager (if not already installed)
curl -LsSf https://astral.sh/uv/install.sh | sh

# 3. Install dependencies
uv sync
uv sync --group dev

# 4. Install Patchright browser
uv run patchright install chromium

# 5. Install pre-commit hooks
uv run pre-commit install

# 6. Create a session (first time only)
uv run -m linkedin_mcp_server --login

# 7. Start the server
uv run -m linkedin_mcp_server

Local Setup Help

CLI Options:

• --login - Open browser to log in and save persistent profile

• --no-headless - Show browser window (useful for debugging scraping issues)

• --log-level {DEBUG,INFO,WARNING,ERROR} - Set logging level (default: WARNING)

• --transport {stdio,streamable-http} - Optional: force transport mode (default: stdio)

• --host HOST - HTTP server host (default: 127.0.0.1)

• --port PORT - HTTP server port (default: 8000)

• --path PATH - HTTP server path (default: /mcp)

• --logout - Clear stored LinkedIn browser profile

• --timeout MS - Browser timeout for page operations in milliseconds (default: 5000)

• --status - Check if current session is valid and exit

• --user-data-dir PATH - Path to persistent browser profile directory (default: ~/.linkedin-mcp/profile)

• --slow-mo MS - Delay between browser actions in milliseconds (default: 0, useful for debugging)

• --user-agent STRING - Custom browser user agent

• --viewport WxH - Browser viewport size (default: 1280x720)

• --chrome-path PATH - Path to Chrome/Chromium executable (for custom browser installations)

• --help - Show help

Note: Most CLI options have environment variable equivalents. See .env.example for details.

HTTP Mode Example (for web-based MCP clients):

uv run -m linkedin_mcp_server --transport streamable-http --host 127.0.0.1 --port 8000 --path /mcp

Live smoke test:

uv run -m linkedin_mcp_server --transport streamable-http --host 127.0.0.1 --port 8080 --path /mcp
uv run python scripts/test_live_tools.py --url http://127.0.0.1:8080/mcp

Target specific tools, add pacing, or retry read_conversation after timeouts:

uv run python scripts/test_live_tools.py \
 --url http://127.0.0.1:8080/mcp \
 --read-only \
 --tool get_conversations \
 --tool read_conversation \
 --read-sleep 3 \
 --read-conversation-retries 2 \
 --retry-backoff-seconds 8 \
 --json-out output/live-smoke.json

Run only the write dry-run tools:

uv run python scripts/test_live_tools.py \
 --url http://127.0.0.1:8080/mcp \
 --write-only

Run a focused read_conversation check against a known thread id:

uv run python scripts/test_live_tools.py \
 --url http://127.0.0.1:8080/mcp \
 --focus-read-conversation \
 --thread-id abc123

Claude Desktop:

{
 "mcpServers": {
 "linkedin": {
 "command": "uv",
 "args": ["--directory", "/path/to/linkedin-agent-mcp", "run", "-m", "linkedin_mcp_server"]
 }
 }
}

stdio is used by default for this config.

Login issues:

• Make sure you have only one active LinkedIn session at a time

• LinkedIn may require a login confirmation in the LinkedIn mobile app for --login

• You might get a captcha challenge if you logged in frequently. The --login command opens a browser where you can solve it manually.

Scraping issues:

• Use --no-headless to see browser actions and debug scraping problems

• Add --log-level DEBUG to see more detailed logging

Session issues:

• Browser profile is stored at ~/.linkedin-mcp/profile/

• Use --logout to clear the profile and start fresh

Python/Patchright issues:

• Check Python version: python --version (should be 3.12+)

• Reinstall Patchright: uv run patchright install chromium

• Reinstall dependencies: uv sync --reinstall

Timeout issues:

• If pages fail to load or elements aren't found, try increasing the timeout: --timeout 10000

• Users on slow connections may need higher values (e.g., 15000-30000ms)

• Can also set via environment variable: TIMEOUT=10000

Custom Chrome path:

• If Chrome is installed in a non-standard location, use --chrome-path /path/to/chrome

• Can also set via environment variable: CHROME_PATH=/path/to/chrome

Feel free to open an issue or PR!

Acknowledgements

Built with FastMCP and Patchright.

⚠️ Use in accordance with LinkedIn's Terms of Service. Web scraping may violate LinkedIn's terms. This tool is for personal use only.

License

This project is licensed under the Apache 2.0 license.

Maintenance

Resources

• GitHub Repository
Need Help?Related Servers