stackcrawler

Stackcrawler MCP for Codex

Configure Stackcrawler’s MCP server in Codex to research website technology stacks from your coding workflow.
Get your API keyExplore the API

1. Set your API key

Get a Stackcrawler API key from your API dashboard. In the terminal where you will launch Codex, set this environment variable with your key:

export STACKCRAWLER_API_KEY="YOUR_STACKCRAWLER_API_KEY"

This shell example is for macOS and Linux. The variable must be available to the Codex process each time you use the connection.

2. Configure the MCP server

Add the following to ~/.codex/config.toml. If a Stackcrawler entry already exists, update it instead of adding a second entry.

[mcp_servers.stackcrawler]
url = "https://api.stackcrawler.com/mcp"
bearer_token_env_var = "STACKCRAWLER_API_KEY"

The configuration references your environment variable so the API key stays out of the config file.

3. Start Codex

Launch a new Codex CLI session from the same terminal and use /mcp to inspect the connection.

See OpenAI’s official MCP documentation for more configuration options.

Try it out

Ask your assistant to use Stackcrawler explicitly:

  • Use Stackcrawler to scan https://example.com and explain its technology stack.
  • Use Stackcrawler to find 10 websites using Shopify with English content.
  • Use Stackcrawler to find websites using the dawn theme.
  • Use Stackcrawler to show websites where Shopify was recently detected as a new technology.

Connection troubleshooting

  • Use the full server URL, including /mcp, and the Streamable HTTP transport.
  • If authentication fails, check your API key or reconnect your Stackcrawler account.
  • Tool calls use your existing API allowance. Check your usage and subscription in the API dashboard if a request is rejected.
  • If browser sign-in cannot complete, contact Stackcrawler support. OAuth must be enabled on the hosted service and linked to your Stackcrawler account.