AskReal.meCONNECT YOUR AI AGENT
Expert knowledge. In your workflow.
Ask a Brain from your AI coding agent. Get advice grounded in real experience, then let your agent put it to work.
How to test your AI usage-based pricing to maximize revenue
Copy one prompt. Your AI Agent handles setup, and you approve access in your browser.
Brain questions use Sparks. Connecting and browsing do not.
Setup instructions
Connect this AI agent to the AskRealMe remote MCP server. Follow the instructions for your client below, use user-level configuration where supported, and preserve existing MCP entries.
Help the user complete browser sign-in, then verify that list_brains and ask_brain are available. Do not spend Sparks on a test question.
For the user's next task, use Brain d46f0d25-33b8-4d57-9ae4-1f0b4d766a68 (How to test your AI usage-based pricing to maximize revenue).
Add the remote server, then sign in to your existing AskRealMe account in the browser. A paid plan is required. AskRealMe hosts the server. Clients with built-in remote MCP support need no extra package; Pi Agent requires the adapter below.
Server URL: https://askrealmeapi-production.up.railway.app/mcp
claude mcp add --transport http --scope user askrealme https://askrealmeapi-production.up.railway.app/mcpOpen /mcp in Claude Code, select AskRealMe, and authenticate. Check that both tools are available.
codex mcp add askrealme --url https://askrealmeapi-production.up.railway.app/mcp
codex mcp login askrealmeIf the tools do not appear in the current conversation, reload its MCP connection or start a new conversation. For longer questions, set tool_timeout_sec = 240 in the AskRealMe entry in ~/.codex/config.toml.
Merge this entry into ~/.cursor/mcp.json, preserving other servers. Then authenticate in Cursor's MCP settings.
{
"mcpServers": {
"askrealme": {
"url": "https://askrealmeapi-production.up.railway.app/mcp"
}
}
}grok mcp add --transport http --scope user askrealme https://askrealmeapi-production.up.railway.app/mcpStart Grok Build and open /mcps. Select AskRealMe and press i to authenticate in your browser. Return to Grok and check that both tools are available. See the Grok Build MCP documentation.
Pi requires an MCP extension. Install the community-maintained pi-mcp-adapter, then restart Pi.
pi install npm:pi-mcp-adapterMerge this entry into ~/.config/mcp/mcp.json, preserving other servers. If you already use a different MCP extension, follow that extension's configuration instructions instead.
{
"mcpServers": {
"askrealme": {
"url": "https://askrealmeapi-production.up.railway.app/mcp",
"auth": "oauth"
}
}
}Restart Pi to load the configuration, run /mcp-auth askrealme, and complete browser sign-in. Use /mcp tools to check that both tools are available. The adapter exposes tools through its mcp proxy by default.
In the IDE, open the agent panel menu → MCP Servers → Manage MCP Servers → View raw config. Merge the entry below into the configuration that opens. In Antigravity CLI, use the global ~/.gemini/config/mcp_config.json file.
{
"mcpServers": {
"askrealme": {
"serverUrl": "https://askrealmeapi-production.up.railway.app/mcp"
}
}
}Use serverUrl, not url. Reload the MCP server and follow the authentication prompt. If needed, open Agent Settings → Customizations and select Authenticate next to AskRealMe. Complete browser sign-in and follow the client's instructions to finish. In the CLI, open /mcp to manage the connection.
See the Antigravity MCP documentation for the configuration location and authentication controls in your version.
Use a Brain
list_brains returns the Top Brains on AskRealMe and spends no Sparks. ask_brain takes a Brain ID, a question, relevant project context, and a new request_id UUID. Successful answers use Sparks. Reuse the same request ID only when retrying the same question.
The returned conversation_id can continue a discussion when useful. Your agent decides how to apply the answer and whether to ask again. Brain knowledge files stay on AskRealMe.
Example: “Use the Loop Engineering Brain to help me apply its approach to this project.”