[Dhruv Shaswat (Dhruv-Sh-git)]

Fixes #9676. This change keeps track of whether a tool call appeared during streaming and uses that information when parsing the final response. I also added a regression test for this case. Track tool calls across Ollama streaming chunks so the final response reports finish_reason='tool_calls' instead of 'stop'. Add a regression test covering a tool call in a non-terminal chunk followed by a terminal chunk without tool calls. This contribution was developed with AI assistance. The changes and tests were reviewed and validated locally.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

[Dhruv Shaswat (Dhruv-Sh-git)]

@microsoft-github-policy-service agree

[Eduard van Valkenburg (eavanvalkenburg)]

/review

[github-actions]

MAF Automated Review — Iteration 1

Result: Findings reported
Scope: full PR (1 commit(s)): f5e11a1ac449
Model: gpt-5.6-sol

Overview

The change correctly keeps tool-call presence request-local, reports the synthesized reason only on the terminal chunk, and retains existing stream cleanup and content-based tool execution. The new regression test covers the reported tool-call-then-stop sequence. However, the same historical flag also overwrites meaningful non-stop terminal reasons such as length, so one Medium contract issue remains.

Reviewed the supplied pull-request change set across correctness, security/reliability, architecture, and failure behavior.
1 verified finding remained after source verification (1 medium) across 1 file. Details are attached to the affected lines below.

Affected areas: python/packages/ollama/agent_framework_ollama/_chat_client.py