microsoft / microsoft/semantic-kernel

.Net: Bug: Semantic Kernel Orchestration: ReasoningEffort setting becomes lost when invoking

Open
#13,935 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

.NET bug triage
Dominant language
C#
Stars
28.6k
Forks
4.8k
Avg merge
14h 13m
Merged PRs (30d)
18

Description

Describe the bug
When involving orchestration, HandOff in my example, ReasoningEffort setting is lost.

To Reproduce
Steps to reproduce the behavior:

  1. Create agent with GPT 5.4 and assign ReasoningEffort as "high"
  2. Create HandOff orchestration using created agent from step 1
  3. Invoke orchestration
  4. Observe HTTP request sent, reasoning_effort field is missing

Expected behavior
When using orchestration, I would expect to maintain reasoning level set on the agent, and that it is not lost during orchestration processing.

Screenshots
Agent HTTP request without orchestration:

Image

Agent HTTP request with orchestration:

Image

Platform

  • Language: C#
  • Source: 1.75.0
  • AI model: gpt-5.4-mini (version:2026-03-17)
  • IDE: Visual Studio 2026 April Release
  • OS: Windows 11

Additional context
See attached sample code.

Program.txt

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the attached Program.txt sample and trace how the agent's ReasoningEffort setting is carried through HandOff orchestration. Compare the outgoing request with and without orchestration, then verify that the request includes the expected reasoning_effort value when the orchestration is invoked.

Written by the indexing model from the issue text.

Assessment

Tech stack
csharp
Domain
ai
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
55/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.