fix(memory): pass agent generation settings to consolidation

[IsamAlBalushi]

Summary

This passes the active agent generation settings through the memory consolidation path so consolidation uses the same temperature, max_tokens, and reasoning_effort as the main agent loop.

Problem

AgentLoop._consolidate_memory() calls MemoryStore.consolidate() without forwarding these settings. As a result, the provider falls back to default values during consolidation.

Fix

• add temperature, max_tokens, and reasoning_effort params to MemoryStore.consolidate()
• forward them from AgentLoop._consolidate_memory()
• pass them into provider.chat_with_retry(...)

Fixes #1823

Notes

This is a minimal change and keeps existing behavior unchanged outside the memory consolidation path.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 976be443f2

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[IsamAlBalushi]

Looks like this was resolved by commit b957dbc which moved generation settings
ownership to the provider layer.

I confirmed locally that memory consolidation now sends the correct
generation parameters. Closing this PR.