[ENHANCEMENT] Roo Code CLI support for OpenAI compatible endpoints

[shruthi-sh]

Problem (one or two sentences)

Currently Roo Code CLI doesn't support OpenAI compatible endpoints . Requesting this as it would be useful for enterprise use cases where the LLM is locally deployed.

Context (who is affected and when)

Users who are using Open AI compatible base url can benefit from this.

Desired behavior (conceptual, not technical)

Login using Open AI base url and api key. Can select model too.

Constraints / preferences (optional)

No response

Request checklist

• I've searched existing Issues and Discussions for duplicates
• This describes a specific problem with clear context and impact

Roo Code Task Links (optional)

No response

Acceptance criteria (optional)

No response

Proposed approach (optional)

No response

Trade-offs / risks (optional)

No response

[roomote-v0]

I've analyzed this issue and here's my proposed implementation plan:

Proposed Plan: Add OpenAI-Compatible Endpoint Support to CLI

The extension already supports an "openai" provider (classified as a CustomProvider) with openAiBaseUrl, openAiApiKey, and openAiModelId settings. The CLI just needs to be wired up to support it.

Changes Required

1. Add "openai" to CLI supported providers (apps/cli/src/types/types.ts)

• Add "openai" to the supportedProviders array so it becomes a valid --provider value

2. Add --base-url CLI flag (apps/cli/src/index.ts)

• Add a new --base-url <url> option for specifying custom API base URLs (e.g. --base-url https://my-llm.example.com/v1)

3. Update FlagOptions type (apps/cli/src/types/types.ts)

• Add baseUrl?: string to FlagOptions

4. Update provider utility (apps/cli/src/lib/utils/provider.ts)

• Add "openai" to envVarMap with OPENAI_API_KEY (distinct from the existing openai-native entry which uses the same env var name -- we'll use OPENAI_COMPATIBLE_API_KEY to avoid conflict, or share the env var if acceptable)
• Update getProviderSettings() to accept an optional baseUrl parameter and set openAiBaseUrl, openAiApiKey, and openAiModelId for the "openai" provider case

5. Update ExtensionHostOptions (apps/cli/src/agent/extension-host.ts)

• Add optional baseUrl?: string field
• Pass it through to getProviderSettings()

6. Wire up in the run command (apps/cli/src/commands/cli/run.ts)

• Read flagOptions.baseUrl and pass it to extensionHostOptions
• Add CliSettings support for persisting a default base URL

7. Add tests

• Update apps/cli/src/lib/utils/__tests__/provider.test.ts for the new provider case
• Update apps/cli/src/commands/cli/__tests__/run.test.ts for the new flag

Usage Examples

# Using OpenAI-compatible endpoint with flags
roo --provider openai --base-url https://my-llm.example.com/v1 --api-key sk-xxx --model my-model "Hello"

# Using environment variables
export OPENAI_COMPATIBLE_API_KEY=sk-xxx
roo --provider openai --base-url https://my-llm.example.com/v1 --model my-model "Hello"

This approach reuses the existing "openai" provider in the extension (which is already designed for OpenAI-compatible endpoints with custom base URLs), so no changes to the core extension are needed.

Please mention @roomote in a comment saying that you approve this plan, and I'll proceed with the implementation.