Skip to content

Feature Request: Different models for coding and testing #146

Description

@jpbester

Currently you can only select a single model for coding and testing. It would be good to be able to select a model for coding and one for testing. For instance, choosing Opus for code, but Sonnet for testing.

Quick Analysis:

  1. In parallel_orchestrator.py (lines 486-503 for coding, 585-595 for testing):
    • Both agent types receive the same self.model parameter
    • The orchestrator passes --model to spawned subprocesses
  2. In client.py (line 251):
    - create_client(model) accepts a single model parameter
  3. No differentiation exists between agent types for model selection

The Fix Needed

To support different models per agent type, you'd need:

  1. Add separate CLI arguments:
    --coding-model opus
    --testing-model sonnet
  2. Update the orchestrator to store both:
    self.coding_model = coding_model
    self.testing_model = testing_model
  3. Pass the appropriate model when spawning:

For coding agents (line 501):

if self.coding_model:
cmd.extend(["--model", self.coding_model])

For testing agents (line 594):

if self.testing_model:
cmd.extend(["--model", self.testing_model])

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions