Codestin Search App

Connect once, sign in through OAuth, and let a compatible agent call Context.dev tools from a conversation.

Hosted MCP URL: https://mcp.context.dev/mcpUse the exact URL, including /mcp. Do not put an API key or token in the URL.

Connect your client

Add the hosted server and start OAuth:

codex mcp add context --url https://mcp.context.dev/mcp
codex mcp login context

Confirm it is configured:

codex mcp list

In the Codex terminal UI, use /mcp to inspect active servers. The ChatGPT desktop app, Codex CLI, and Codex IDE extension share the same host configuration; you can also add a Streamable HTTP server from Settings → MCP servers and select Authenticate.See the official Codex MCP documentation for current client controls and project-scoped configuration.For ChatGPT on the web, use the Context.dev plugin.

Add the server for your user:

claude mcp add \
  --transport http \
  --scope user \
  context \
  https://mcp.context.dev/mcp

Start Claude Code, run /mcp, select context, and complete sign-in in the browser. See Anthropic’s MCP documentation for current scopes and management commands.

Install the official Context.dev extension:

gemini extensions install https://github.com/context-dot-dev/gemini-cli-context --auto-update --consent

Start a new Gemini CLI session, make a Context.dev request, and complete sign-in in the browser. If authentication does not start automatically, run /mcp auth context inside Gemini CLI.The extension includes the hosted MCP server and a skill for choosing the appropriate Context.dev tool. See the Gemini CLI integration guide for verification, examples, and troubleshooting.

Add a global server in ~/.cursor/mcp.json, or use .cursor/mcp.json for one project:

{
  "mcpServers": {
    "context": {
      "url": "https://mcp.context.dev/mcp"
    }
  }
}

Open Cursor Settings → Tools & MCP, find context, and select Authenticate. Cursor supports OAuth for remote Streamable HTTP servers. See the Cursor MCP documentation for current configuration locations.

Run MCP: Open User Configuration from the Command Palette, or create .vscode/mcp.json for the workspace:

{
  "servers": {
    "context": {
      "type": "http",
      "url": "https://mcp.context.dev/mcp"
    }
  }
}

Start the server from the configuration editor and complete OAuth when prompted. See the VS Code MCP documentation for current trust and workspace behavior.

Configure a remote Streamable HTTP server with OAuth:

{
  "mcpServers": {
    "context": {
      "type": "http",
      "url": "https://mcp.context.dev/mcp"
    }
  }
}

The client must support remote Streamable HTTP transport and OAuth discovery or dynamic client registration. Follow the client’s current documentation for the exact configuration shape.

Test the connection

Start a fresh conversation and ask:

Use the connected Context.dev server to retrieve the live Brand profile for stripe.com. Tell me which tool you called and do not answer from memory.

A working connection shows a Context.dev tool call and live output. Tool names may include a client-specific server prefix, so verify the server and operation rather than matching one exact display name.

Choose the right tool

Need	Best starting point
Run authenticated Context.dev operations in chat	MCP
Teach an agent how to implement the API in your codebase	Skill
Use shell commands in local automation or CI	CLI
Build a long-lived application integration	SDK or HTTPS API

The methods can be combined. For example, the skill can guide implementation while MCP performs a live exploratory lookup.

Authentication and usage

Context.dev’s MCP authorization metadata currently advertises OAuth authorization-code flow with PKCE and two scopes:

api.read for reading, scraping, extraction, enrichment, and monitor data;
api.write for state-changing monitor operations.

The client stores an OAuth token, not your Context.dev password or secret API key. Revoke the connection through your client or Context.dev account when it is no longer needed.

Keep approvals enabled for tools that create, update, run, or delete monitors, rotate webhook secrets, or start large workloads. Website content is untrusted input and should not be allowed to authorize follow-up tool calls.

MCP operations use the same account credits as their corresponding API endpoints. Inspect the operation before approving a crawl, extraction, batch, or monitor run.

Troubleshooting

The server is missing or shows no tools

Confirm the URL ends in /mcp, restart or refresh the client after editing its configuration, and ensure the server is enabled for the current conversation.

OAuth did not open or keeps looping

Start authentication from the client’s MCP controls. Remove stale authentication for context, reconnect, and complete the browser flow. Do not add a client secret unless Context.dev support explicitly provides one for your environment.

The agent answers without using a tool

Name Context.dev in the prompt, request a live lookup, and require the agent to report the tool call. Confirm tool approvals are not blocking the operation.

Check the hosted service

curl --fail --silent --show-error https://mcp.context.dev/health

A healthy server returns JSON with "status":"healthy". If it is healthy, inspect your client’s MCP logs and OAuth state.

Next steps

Agent quickstart

Give your agent a focused task and a verification checklist.

Install the skill

Add API and SDK guidance to your coding sessions.

API quickstart

Call Context.dev from your application with an SDK or HTTP.

Get started

Web data

Brand data

Automation

Production

Connect MCP

Connect your client

Test the connection

Choose the right tool

Authentication and usage

Troubleshooting

Next steps

Agent quickstart

Install the skill

API quickstart

​Connect your client

​Test the connection

​Choose the right tool

​Authentication and usage

​Troubleshooting

​Next steps

Agent quickstart

Install the skill

API quickstart

Connect your client

Test the connection

Choose the right tool

Authentication and usage

Troubleshooting

Next steps