Framework adapters

Framework adapters are explicit output formats, not separate Authlane gateways. Your backend scopes the request with externalUserId; the resulting handlers run beside your agent and call providers directly.

TypeScript

import { Authlane } from '@authlane/sdk';
import { mastraAI } from '@authlane/ai/mastra';

const authlane = new Authlane({
  apiKey: process.env.AUTHLANE_API_KEY!,
  baseUrl: 'https://app.authlane.io',
});

const { data: tools, error } = await authlane
  .user('user_123')
  .tools.list({ adapter: mastraAI() });

Choose the adapter at the integration boundary:

Framework	Import	Adapter
Vercel AI SDK	@authlane/ai/vercel	vercelAI()
OpenAI Agents	@authlane/ai/openai	openAIAgents()
Mastra	@authlane/ai/mastra	mastraAI()
Local MCP server	@authlane/ai/mcp	mcpServer()

Python

import os

from authlane import Authlane
from authlane.adapters import agno, generic, langchain, openai_agents

with Authlane(api_key=os.environ["AUTHLANE_API_KEY"]) as authlane:
    user = authlane.user("user_123")
    agno_tools = user.tools.list(adapter=agno())
    langchain_tools = user.tools.list(adapter=langchain())
    openai_tools = user.tools.list(adapter=openai_agents())
    portable_tools = user.tools.list(adapter=generic())

Identity and execution boundary

The user_123 value must come from your authenticated SaaS session. Never accept an arbitrary external user ID from an untrusted client. Authlane returns definitions and scoped local handlers; it does not receive tool calls and does not expose a hosted MCP server.