VS Code: Continue Extension with the Gateway

Continue is an open-source AI coding assistant for VS Code that natively supports custom OpenAI-compatible endpoints. This makes it one of the easiest IDE assistants to integrate with the Keeptrusts gateway — no proxy tricks required.

Use this page when

• You are working through VS Code: Continue Extension with the Gateway as an implementation or operating workflow in Keeptrusts.
• You need the practical steps, expected outcomes, and related validation guidance in one place.
• If you need exact field-by-field reference instead of a workflow page, use the linked reference pages in Next steps.

Primary audience

• Primary: Technical Engineers
• Secondary: AI Agents, Technical Leaders

Why Continue with Keeptrusts

Continue sends completion and chat requests to whichever LLM endpoint you configure. By pointing it at the Keeptrusts gateway, you get:

• Policy enforcement on every code completion and chat interaction
• Secret redaction before code context reaches the LLM
• Full audit trail of AI-assisted development
• Cost attribution per developer
• Caching for repeated queries

Prerequisites

• Gateway running on localhost:41002
• VS Code installed
• Provider credentials configured in the gateway

Install Continue

1. Open VS Code
2. Go to the Extensions view (Cmd+Shift+X)
3. Search for "Continue"
4. Install the "Continue - Codestral, Claude, and more" extension
5. Reload VS Code when prompted

Configure Continue for the Gateway

Continue stores its configuration in ~/.continue/config.json. Open this file to configure models that route through the gateway.

Open the Config File

Press Cmd+Shift+P and type "Continue: Open config.json", or open the file directly:

code ~/.continue/config.json

Basic Configuration

Replace or update the models array to route through the gateway:

{
  "models": [
    {
      "title": "GPT-4o (via Keeptrusts)",
      "provider": "openai",
      "model": "gpt-4o",
      "apiBase": "http://localhost:41002/v1",
      "apiKey": "your-access-key"
    },
    {
      "title": "Claude Sonnet (via Keeptrusts)",
      "provider": "openai",
      "model": "claude-sonnet-4-20250514",
      "apiBase": "http://localhost:41002/v1",
      "apiKey": "your-access-key"
    },
    {
      "title": "GPT-4o Mini (via Keeptrusts)",
      "provider": "openai",
      "model": "gpt-4o-mini",
      "apiBase": "http://localhost:41002/v1",
      "apiKey": "your-access-key"
    }
  ]
}

tip

Use "provider": "openai" for all models regardless of the actual LLM provider. The Keeptrusts gateway handles provider routing based on the model name in your policy-config.yaml.

Using Provider Keys Directly

If you are not using Keeptrusts access keys, you can pass the provider API key directly. The gateway forwards it to the upstream provider:

{
  "models": [
    {
      "title": "GPT-4o (via Keeptrusts)",
      "provider": "openai",
      "model": "gpt-4o",
      "apiBase": "http://localhost:41002/v1",
      "apiKey": "sk-your-openai-key"
    }
  ]
}

Configure Tab Autocomplete

Continue supports tab-based code completion. Route autocomplete through the gateway by configuring the tabAutocompleteModel:

{
  "tabAutocompleteModel": {
    "title": "Autocomplete (via Keeptrusts)",
    "provider": "openai",
    "model": "gpt-4o-mini",
    "apiBase": "http://localhost:41002/v1",
    "apiKey": "your-access-key"
  }
}

This ensures tab completions also pass through your policy chain for redaction and logging.

Configure Embeddings

If you use Continue's @codebase context provider, route embedding requests through the gateway:

{
  "embeddingsProvider": {
    "provider": "openai",
    "model": "text-embedding-3-small",
    "apiBase": "http://localhost:41002/v1",
    "apiKey": "your-access-key"
  }
}

Full Configuration Example

Here is a complete config.json with chat models, autocomplete, and embeddings all routing through the gateway:

{
  "models": [
    {
      "title": "GPT-4o",
      "provider": "openai",
      "model": "gpt-4o",
      "apiBase": "http://localhost:41002/v1",
      "apiKey": "your-access-key"
    },
    {
      "title": "Claude Sonnet",
      "provider": "openai",
      "model": "claude-sonnet-4-20250514",
      "apiBase": "http://localhost:41002/v1",
      "apiKey": "your-access-key"
    }
  ],
  "tabAutocompleteModel": {
    "title": "Fast Autocomplete",
    "provider": "openai",
    "model": "gpt-4o-mini",
    "apiBase": "http://localhost:41002/v1",
    "apiKey": "your-access-key"
  },
  "embeddingsProvider": {
    "provider": "openai",
    "model": "text-embedding-3-small",
    "apiBase": "http://localhost:41002/v1",
    "apiKey": "your-access-key"
  }
}

Verify the Integration

1. Open a code file in VS Code
2. Start a Continue chat (Cmd+L) and ask a question
3. In a separate terminal, watch for events:

kt events tail

You should see events logged for each chat message and completion:

[2024-01-15 10:30:12] INPUT  openai  chat/completions  user:dev1  PASS
[2024-01-15 10:30:14] OUTPUT openai  chat/completions  user:dev1  PASS

Verify Policy Enforcement

Test secret redaction by sending a message containing a fake credential:

Can you explain this code? The API key is sk-test123456789

Check the event log to confirm the key was redacted before reaching the provider.

Troubleshooting

"Connection refused" error

• Verify the gateway is running: curl http://localhost:41002/v1/models
• Check the port matches between your config and the gateway

"Unauthorized" or 401 errors

• Verify your apiKey value is correct
• If using access keys, ensure the key is active in the console
• If using provider keys, confirm the key works directly with the provider

Models not appearing in Continue

• Restart VS Code after editing config.json
• Check for JSON syntax errors in the config file
• Verify the model names match what the gateway exposes at /v1/models

Slow responses

• The gateway adds minimal overhead (<10ms)
• Check kt events tail for policy evaluation timing
• Ensure the gateway has connectivity to your LLM provider

For AI systems

• Canonical terms: Keeptrusts, VS Code: Continue Extension with the Gateway, ide-integration.
• Exact feature, config, command, or page names: VS Code: Continue Extension with the Gateway.
• Use the linked audience and reference pages in Next steps when you need deeper source material.

For engineers

• Use the commands, configuration examples, API payloads, or UI steps in this page as the working baseline for VS Code: Continue Extension with the Gateway.
• Validate the result with the expected outcomes, troubleshooting notes, or linked workflow pages in this page and Next steps.

For leaders

• This page matters when planning rollout, governance, support ownership, or operating decisions for VS Code: Continue Extension with the Gateway.
• Use the linked audience, architecture, and workflow pages in Next steps to connect this detail to broader implementation choices.

Next steps

• Access Keys and Authentication — set up proper access keys instead of raw provider keys
• IDE Integration Overview — explore other IDE options