Skip to content

Carry effort and extended-thinking settings into routed LLM configs - #105

Merged
knight22-21 merged 1 commit into
knight22-21:mainfrom
mayuriphad:fix/router-inherits-effort-thinking
Oct 13, 2026
Merged

knight22-21 merged 1 commit into
knight22-21:mainfrom
mayuriphad:fix/router-inherits-effort-thinking

Conversation

@mayuriphad

Copy link
Copy Markdown
Contributor

Summary

  • get_llm_for_task() built the routed LLMConfig without effort, extended_thinking or thinking_budget_tokens, so every routed call (planning, coding, reviewing, cheap) used the defaults (effort="high", thinking off) whatever the user set under [llm].
  • Copy the three fields from the base config.llm.
  • Add a regression test; it fails on main and passes with the change.

Testing

  • pytest tests/test_phase5.py: 29 passed (28 passed + 1 failed without the fix).
  • ruff check on the changed files passes.

Closes #100

get_llm_for_task built a fresh LLMConfig without effort, extended_thinking or thinking_budget_tokens, so routed calls always used the defaults regardless of [llm] settings.

Signed-off-by: mayuriphad <mayuriphad656@gmail.com>
Copilot AI balanced review requested due to automatic review settings October 13, 2026 03:57

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@knight22-21
knight22-21 merged commit cd55fc2 into knight22-21:main Oct 13, 2026
5 checks passed
@knight22-21

Copy link
Copy Markdown
Owner

Thanks for the fix, @mayuriphad! Forwarding \effort, \extended_thinking, and \ hinking_budget_tokens\ into the routed LLM configs ensures user settings aren't silently dropped during task routing.

If you find DevAgent useful, a ⭐ on the repo would mean a lot — no pressure though!

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

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

get_llm_for_task() drops effort and extended thinking parameters from base config

3 participants