VS Code - Timely.ai

VS Code does not have native support for external documentation sources like Cursor, but two extensions cover this use case in different ways: Continue.dev (semantic access to documentation) and GitHub Copilot (via manual context comments).

Continue.dev

Continue.dev is an open-source extension that adds an AI chat to VS Code with support for context providers, including @docs. With it, you index the Timely.ai documentation and mention it in prompts the same way you would in Cursor.

Setup

Install the extension

Open VS Code, go to Extensions (Cmd+Shift+X), and search for Continue. Install the official extension.

Open config.json

After installation, click the Continue icon in the sidebar. In the panel, click the gear icon or open the file directly:

~/.continue/config.json

Add Timely.ai as docs

In config.json, add the entry inside "docs":

{
  "docs": [
    {
      "title": "Timely.ai",
      "startUrl": "https://docs.timelyai.com.br",
      "rootUrl": "https://docs.timelyai.com.br",
      "maxDepth": 3
    }
  ]
}

Save the file. Continue will start indexing automatically.

Use in prompts

In the Continue panel, mention @docs and select Timely.ai:

@docs How do I create a contact via API and associate it with a conversation?

Complete config.json (minimal example)

{
  "models": [
    {
      "title": "Claude 3.5 Sonnet",
      "provider": "anthropic",
      "model": "claude-3-5-sonnet-20241022",
      "apiKey": "your-anthropic-key"
    }
  ],
  "docs": [
    {
      "title": "Timely.ai",
      "startUrl": "https://docs.timelyai.com.br",
      "rootUrl": "https://docs.timelyai.com.br",
      "maxDepth": 3
    }
  ],
  "contextProviders": [
    { "name": "code" },
    { "name": "docs" },
    { "name": "diff" }
  ]
}

Replace "your-anthropic-key" with your API key or configure another provider of your choice. Continue supports OpenAI, Anthropic, Ollama, and others.

GitHub Copilot

GitHub Copilot does not have native support for context providers from external URLs. To give it context about the Timely.ai API, use comments at the top of files where you will be working on the integration.

Strategy with context comments

Add a comment block at the top of the file before asking Copilot for suggestions:

/**
 * Timely.ai API Integration
 * Base URL: https://api.timelyai.com.br
 * Auth: x-api-key header with your API key
 * Docs: https://docs.timelyai.com.br
 *
 * Endpoints used in this file:
 * - GET  /v1/contacts         → list contacts
 * - POST /v1/contacts         → create contact (name, phone required)
 * - POST /v1/conversations/{id}/messages → send message
 *
 * Rate limit: 60 req/min per workspace
 * Pagination: page and limit parameters, response includes meta.total
 */

With this block present, Copilot generates suggestions that are more aligned with the real API structure.

Alternative: use llms.txt as an inline reference

For longer sessions, download the llms.txt and open it as a file in the editor. Copilot treats open files in the window as additional context in some configurations.

curl https://docs.timelyai.com.br/llms.txt -o timely-llms.txt

Open the file in VS Code and keep it in a tab while working on the integration files.

Quick comparison

Feature	Continue.dev	GitHub Copilot
Automatic docs indexing	Yes	No
Mention `@docs` in prompt	Yes	No
Inline editor suggestions	Yes	Yes
Cost	Free	GitHub subscription
Supported models	Many	OpenAI (via GitHub)

Skills

​Continue.dev

​Setup

​Complete config.json (minimal example)

​GitHub Copilot

​Strategy with context comments

​Alternative: use llms.txt as an inline reference

​Quick comparison

Continue.dev

Setup

Complete config.json (minimal example)

GitHub Copilot

Strategy with context comments

Alternative: use llms.txt as an inline reference

Quick comparison