Cline

Cline CLI accepts openai as shorthand for its OpenAI-compatible provider and stores it as openai-compatible. Configure it with the Pool API key, the Codex Pooler /v1 base URL, and the model id that your assigned Pool can serve.

cline auth \
  --provider openai \
  --apikey "$CODEX_POOLER_API_KEY" \
  --baseurl https://codex-pooler.example.com/v1 \
  --modelid gpt-5.6-terra

For local setup, change --baseurl to http://localhost:4000/v1.

Cline’s user-facing model metadata names are contextWindow, maxInputTokens, and maxTokens. If you add a manual Codex Pooler model entry in Cline settings, use contextWindow: 258400, maxInputTokens: 258400, and maxTokens: 128000 for gpt-5.6-terra. Cline applies its fixed 90% compaction ratio to the effective maximum input, so it starts compaction at 232560 tokens. maxTokens remains a separate response cap and does not move the local compaction boundary. When available, /v1/models.context_length is the authoritative effective value for contextWindow.

Check the headless CLI path after saving auth:

cline --provider openai \
  --model gpt-5.6-terra \
  --json \
  --auto-approve false \
  'Reply with exactly: cline ok'

For optional operator MCP in Cline CLI, add the remote server to ~/.cline/mcp.json. Codex Pooler does not require this for model use. The VS Code extension opens its own MCP settings JSON from the Cline MCP Servers panel; use the same mcpServers shape there.

{
  "mcpServers": {
    "codex_pooler": {
      "url": "https://codex-pooler.example.com/mcp",
      "headers": {
        "Authorization": "Bearer <operator-mcp-token>"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

For local MCP setup, change the MCP url to http://localhost:4000/mcp.

Use a Pool API key for /v1 model requests and an operator MCP token for /mcp. Do not reuse the Pool API key for MCP.