Continue Adapter

Continue is an open-source AI coding assistant that works in VS Code and JetBrains. The FORG Continue adapter uses Continue's context provider API to emit signals after each AI interaction.

Prerequisites

• Continue extension v0.9.0 or later
• FORG CLI installed and activated

Installation

Add the FORG context provider to your Continue config at ~/.continue/config.json:

{
  "models": [
    {
      "title": "Claude Opus",
      "provider": "anthropic",
      "model": "claude-opus-4-5",
      "apiKey": "sk-ant-..."
    }
  ],
  "contextProviders": [
    {
      "name": "forg",
      "params": {
        "agentPort": 6247,
        "projectId": "proj_my-continue-work"
      }
    }
  ]
}

The FORG context provider is built into Continue v0.9.0+. If you're on an older version, install it manually:

npm install -g @forg-pro/continue-provider

What gets captured

• All Continue chat interactions
• Edit and refactor requests
• Model used, token counts, latency, and cost
• Session boundaries (new session after 30 min inactivity)

Supported model providers

The FORG adapter captures signals regardless of which model provider Continue is configured to use: Anthropic, OpenAI, Ollama, Azure OpenAI, Mistral, Cohere, etc. Token counts for locally-hosted models (Ollama) are estimated.

Verification

# In Continue, open the chat panel and ask something.
# Then check:
forg status --adapter
# Last signal: 5s ago
# Model: claude-opus-4-5

Troubleshooting