# Using environments with the Vercel AI SDK

[Environments](/docs/mcp/secure/environments) in Gram manage authentication credentials, API keys, and configuration variables for [toolsets](/docs/mcp/reference/concepts/toolsets). When building agent applications with the Vercel AI SDK or other AI frameworks, environments provide a secure way to manage the credentials needed to execute tools.

## Overview

Environments enable secure credential management without hardcoding sensitive information. They work by:

- Storing credentials and configuration in Gram's secure environment system
- Binding tools to a specific environment at initialization
- Executing all tool calls using the credentials from that environment

This approach is particularly useful for:

- **Multi-environment deployments**: Separate development, staging, and production credentials
- **Team collaboration**: Share toolsets without exposing credentials
- **Enterprise security**: Centralized credential management and rotation

Learn more about [environment concepts](/docs/mcp/secure/environments) and how to [configure environments](/docs/mcp/build/gram-functions/configuring-environments).

## Using environments with Vercel AI SDK

The Vercel AI SDK integration demonstrates how to bind tools to a specific environment. The `environment` parameter in the `tools()` method determines which credentials the tools will use at runtime.

```ts filename="vercel-example.ts" {15}

const key = process.env.GRAM_API_KEY;
const vercelAdapter = new VercelAdapter({ apiKey: key });

const openai = createOpenAI({
  apiKey: process.env.OPENAI_API_KEY,
});

const tools = await vercelAdapter.tools({
  project: "acme",
  toolset: "marketing",
  environment: "production",
});

const result = await generateText({
  model: openai("gpt-4"),
  tools,
  maxSteps: 5,
  prompt: "Can you tell me what tools you have available?",
});

console.log(result.text);
```

In this example, all tools in the `marketing` toolset execute using credentials from the `production` environment. This keeps production API keys secure and separate from development credentials.

See the [Vercel AI SDK integration guide](/docs/mcp/build/integrate/api-clients/using-vercel-ai-sdk-with-gram-mcp-servers) for complete setup instructions.

## Using environments with OpenAI Agents SDK

The OpenAI Agents SDK follows the same pattern, binding tools to an environment at initialization:

```py filename="openai-agents-example.py" {17}

const vercelAdapter = new VercelAdapter({
  apiKey: process.env.GRAM_API_KEY,
  environmentVariables: {
    ACME_API_KEY: process.env.ACME_API_KEY,
    ACME_SERVER_URL: "https://janesmith.acme.com",
  },
});
```

The `environmentVariables` object accepts any key-value pairs that the toolset's functions require. This approach bypasses Gram's environment system entirely, using credentials provided at runtime instead.

### Python example

```py filename="byo-env-vars.py" {6-9}
import os
from gram_ai.openai_agents import GramOpenAIAgents

gram = GramOpenAIAgents(
    api_key=os.getenv("GRAM_API_KEY"),
    environment_variables= {
        "ACME_API_KEY": os.getenv("ACME_API_KEY"),
        "ACME_SERVER_URL": "https://janesmith.acme.com",
    }
)
```

Both SDK adapters support the same `environment_variables` parameter, providing flexibility in how credentials are managed.

## Choosing between environments and direct variables

Consider these factors when deciding between managed environments and direct variables:

**Use managed environments when:**
- Deploying applications with fixed credentials
- Managing multiple environments (dev, staging, production)
- Sharing toolsets across teams without exposing credentials
- Implementing centralized credential rotation

**Use direct variables when:**
- Building multi-tenant applications
- Creating end-user tools where users supply credentials
- Testing with dynamic credential sets
- Implementing user-specific customizations

Both approaches can be combined in the same application for different use cases.

## Next steps

- Learn about [environment concepts and management](/docs/mcp/secure/environments)
- Explore [other AI framework integrations](/docs/mcp/build/integrate/api-clients/using-anthropic-api-with-gram-mcp-servers)
- Set up [API keys](/docs/mcp/secure/api-keys) for Gram SDK access
- Read about [toolset organization](/docs/mcp/reference/concepts/toolsets)