[FEATURE] Model alias for sonnet default -> [1m] after filling context window

Preflight Checklist

• [x] I have searched existing requests and this feature hasn't been requested yet
• [x] This is a single feature request (not multiple features)

Problem Statement

I often don't know at the beginning of a task whether I'll need Sonnet[1m] or Default sonnet's context window limit will be sufficient. I have 3 choices:

1. Choose Default and allow context window to be compacted, potentially losing context
2. Choose Sonnet[1m] and pay more from the beginning, whether I end up needing it or not
3. Choose Default and switch to Sonnet[1m] if I see the warning indicating that context size is running low.

The workflow for # 3 suffers from 2 problems:

1. Requires me to cancel mid execution to switch models, then type 'continue' after
2. If I'm letting the agent run on auto-accept edit mode, or with --dangerously-skip-permissions, without actively monitoring it, I might miss the context size warning and miss my opportunity to switch models before context is compacted.

Proposed Solution

Create a new model alias which automatically switches models when available context runs low or out.

Alternative Solutions

When context becomes full, prompt user with a choice for either switching models or compacting the window

Priority

Low - Nice to have

Feature Category

Configuration and settings

Use Case Example

Example scenario:

1. I have a medium+ sized task with important, nuanced context required to execute successfully
2. I don't know if the entire task can be completed before the Default model's context window fills up
3. I'm worried about losing important context when it is compacted.
4. I don't want to have to spend the money on Sonnet[1m] from the beginning, if I'm not sure I'm going to need it.

Additional Context

_No response_