agentbay_brain_setup

Instructions

Implementation Reference

• src/index.ts:976-994 (handler)

  The handler function for agentbay_brain_setup tool. It calls the /api/v1/brain/setup endpoint with name, description, framework, and model parameters, then returns project ID, agent ID, brain slug, and OpenClaw config if available.

    async ({ name: brainName, description: brainDesc, framework: fw, model: mdl }) => {
      const data = await apiPost('/api/v1/brain/setup', {
        name: brainName,
        description: brainDesc,
        framework: fw,
        model: mdl,
      });
      if (data.error) return { content: [{ type: 'text' as const, text: `Error: ${data.error}` }] };
  
      let text = `Knowledge Brain created!\n\nProject ID: ${data.projectId}\nAgent ID: ${data.agentId}\nSlug: ${data.brainSlug}`;
      if (data.openclawConfig) {
        if (data.openclawConfig.claudeCommand) text += `\n\nOpenClaw setup:\n${data.openclawConfig.claudeCommand}`;
        if (data.openclawConfig.soulMdAddition) text += `\n\nSOUL.md addition:\n${data.openclawConfig.soulMdAddition}`;
        if (data.openclawConfig.bootstrapMd) text += `\n\nBOOTSTRAP.md:\n${data.openclawConfig.bootstrapMd}`;
      }
      text += `\n\nNext: Import knowledge with agentbay_brain_import`;
      return { content: [{ type: 'text' as const, text }] };
    }
  );

• src/index.ts:970-975 (schema)

  Input schema for agentbay_brain_setup using Zod. Accepts: name (string, required), description (optional string), framework (optional string), model (optional string).

  {
    name: z.string().describe('Agent name (e.g., "Moonsa", "my-agent")'),
    description: z.string().optional().describe('Brain description'),
    framework: z.string().optional().describe('Agent framework (openclaw, langchain, crewai, custom)'),
    model: z.string().optional().describe('Primary model the agent uses'),
  },

• src/index.ts:967-994 (registration)

  Registration of the 'agentbay_brain_setup' tool on the MCP server via server.tool(), with its description, schema, and handler function.

  server.tool(
    'agentbay_brain_setup',
    'Create a Knowledge Brain for your agent in one call. Returns project ID, agent ID, and all configs needed to connect.',
    {
      name: z.string().describe('Agent name (e.g., "Moonsa", "my-agent")'),
      description: z.string().optional().describe('Brain description'),
      framework: z.string().optional().describe('Agent framework (openclaw, langchain, crewai, custom)'),
      model: z.string().optional().describe('Primary model the agent uses'),
    },
    async ({ name: brainName, description: brainDesc, framework: fw, model: mdl }) => {
      const data = await apiPost('/api/v1/brain/setup', {
        name: brainName,
        description: brainDesc,
        framework: fw,
        model: mdl,
      });
      if (data.error) return { content: [{ type: 'text' as const, text: `Error: ${data.error}` }] };
  
      let text = `Knowledge Brain created!\n\nProject ID: ${data.projectId}\nAgent ID: ${data.agentId}\nSlug: ${data.brainSlug}`;
      if (data.openclawConfig) {
        if (data.openclawConfig.claudeCommand) text += `\n\nOpenClaw setup:\n${data.openclawConfig.claudeCommand}`;
        if (data.openclawConfig.soulMdAddition) text += `\n\nSOUL.md addition:\n${data.openclawConfig.soulMdAddition}`;
        if (data.openclawConfig.bootstrapMd) text += `\n\nBOOTSTRAP.md:\n${data.openclawConfig.bootstrapMd}`;
      }
      text += `\n\nNext: Import knowledge with agentbay_brain_import`;
      return { content: [{ type: 'text' as const, text }] };
    }
  );

• src/index.ts:947-964 (helper)

  Related tool 'agentbay_agent_register' (alias) which also calls /api/v1/brain/setup but with a slightly different parameter set (no model), using a simpler response format.

  server.tool(
    'agentbay_agent_register',
    'Register this agent with AgentBay. Required before using agent memory tools (agent_memory_record, agent_memory_query). Creates a Knowledge Brain and links it to your API key.',
    {
      name: z.string().describe('Agent name (e.g., "codex", "my-agent")'),
      description: z.string().optional().describe('Agent description'),
      framework: z.string().optional().describe('Agent framework (codex, openclaw, langchain, crewai, custom)'),
    },
    async ({ name: brainName, description: brainDesc, framework: fw }) => {
      const data = await apiPost('/api/v1/brain/setup', {
        name: brainName,
        description: brainDesc,
        framework: fw,
      });
      if (data.error) return { content: [{ type: 'text' as const, text: `Error: ${data.error}` }] };
      return { content: [{ type: 'text' as const, text: `Agent registered!\n\nAgent ID: ${data.agentId}\nProject ID: ${data.projectId}\n\nYou can now use agentbay_agent_memory_record and agentbay_agent_memory_query.` }] };
    }
  );