[Hashim Khan (Hashim1999164)]

Summary

InMemoryResponseProvider.get_history_item_ids and FileResponseProvider.get_history_item_ids truncated with resolved[:limit], which kept the oldest IDs and dropped recent turns. Conversation history should keep the newest N items.

This changes both providers to resolved[-limit:] and tightens the unit test to assert the newest IDs are retained.

Fixes #48514

Changelog

Updated sdk/agentserver/azure-ai-agentserver-responses/CHANGELOG.md under 2.1.0b2 (Unreleased).

Test plan

1. pytest sdk/agentserver/azure-ai-agentserver-responses/tests/unit/test_in_memory_provider_crud.py::test_history__respects_limit
2. Confirm truncated results are the last N chronological IDs

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).
7 pipeline(s) were filtered out due to trigger conditions.
There may be pipelines that require an authorized user to comment /azp run to run.

[github-actions]

Thank you for your contribution Hashim Khan (@Hashim1999164)! We will review the pull request and get back to you soon.

[Copilot]

Pull request overview

Updates history truncation to retain the newest items.

Changes:

• Uses tail slicing in memory and file stores.
• Strengthens the in-memory limit test.
• Adds a changelog entry.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
_memory.py	Keeps newest in-memory history IDs.
_file.py	Keeps newest file-backed history IDs.
test_in_memory_provider_crud.py	Verifies newest IDs are retained.
CHANGELOG.md	Documents the bug fix.

---

💡 Add a code-review agent skill for context-aware, tailored reviews. Learn more in the docs.

[Copilot]

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated no new comments.

Suppressed comments (1)

sdk/agentserver/azure-ai-agentserver-responses/CHANGELOG.md:7

• The PR description says this fix belongs to 2.1.0b2 (Unreleased), but this instead changes the existing 2.1.0b1 release date and attributes the fix to that release. Add a new unreleased section and preserve the prior 2.1.0b1 (2026-08-11) heading so users can identify which package version contains the fix.

## 2.1.0b1 (2026-08-13)

### Bugs Fixed

- `get_history_item_ids` on the in-memory and file stores now keeps the newest item IDs when applying `limit`. (#48514)

[Hashim Khan (Hashim1999164)]

Pushed a changelog follow up.

test_file_response_store_parity.py already expected the newest IDs (["out1", "out2", "out3"]), so that test was left unchanged.

The changelog now keeps 2.1.0b1 (2026-08-11) as released and records the history limit fix under 2.1.0b2 (Unreleased), naming FileResponseStore and InMemoryResponseProvider.

[Copilot]

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated no new comments.

Suppressed comments (2)

sdk/agentserver/azure-ai-agentserver-responses/azure/ai/agentserver/responses/store/_memory.py:329

• The function docstring for get_history_item_ids still describes returning IDs “up to limit” without specifying truncation semantics. Since this change makes limit mean “keep the most recent N”, please update the docstring (and any related protocol/interface docs, if present) to explicitly state that limit retains the newest IDs while preserving chronological order within the retained slice.

            # Keep the most recent N history items (drop oldest when over limit).
            return resolved[-limit:]

sdk/agentserver/azure-ai-agentserver-responses/azure/ai/agentserver/responses/store/_file.py:583

• Same as the in-memory provider: please update the get_history_item_ids documentation for the file-backed store to clarify that limit keeps the most recent N IDs (drops oldest), so consumers don’t assume the previous “first N” behavior.

            # Keep the most recent N history items (drop oldest when over limit).
            return resolved[-limit:]

[github-actions]

[Pilot] PR Pipeline Failure Analysis

A CI pipeline failed on this pull request. Here is an automated analysis of what went wrong and how to get the build green.

What failed

The python - pullrequest (Build Analyze) stage failed (build #6707528) during changelog validation for azure-ai-agentserver-responses. Two validation errors were reported:

1. Invalid date: The changelog entry for 2.1.0b1 is dated 2026-08-11, but the validator requires it to be the latest date in the file.
2. Empty section: The Bugs Fixed section in the changelog entry has no content — either remove the empty section header or add content to it.

Recommended next steps

• In sdk/agentserver/azure-ai-agentserver-responses/CHANGELOG.md, update the date on the 2.1.0b1 entry to today's date (or the intended release date, ensuring it is later than all other dates in the file).
• Remove the empty ### Bugs Fixed section from the 2.1.0b1 entry, or add actual content to it.
• See the CI troubleshooting guide: https://aka.ms/ci-fix
• Push new commits to address the failures; this comment updates automatically on the next failing run.

Raw pipeline analysis (azsdk ci analyze)

Analyzing pipeline https://github.com/Azure/azure-sdk-for-python/pull/48560...
Getting failed workflow runs for commit 6220bba239d1621e4f4363dd2001a27d8c6fece1 in Azure/azure-sdk-for-python
Build: 6707528 Project: public PipelineUrl: https://dev.azure.com/azure-sdk/public/_build/results?buildId=6707528

Failed Tasks

### Errors:
Verifying changelog for azure-ai-agentserver-responses
Found the following change log entry for version '2.1.0b1' in [sdk/agentserver/azure-ai-agentserver-responses/CHANGELOG.md].

## 2.1.0b1 (2026-08-11)
...
[debug]Verifying as a release build because the changelog entry has a valid date.
##[error]Invalid date [ 2026-08-11 ]. The date for the changelog being released must be the latest in the file.
##[error]The changelog entry has the following sections with no content (Bugs Fixed). Please ensure to either remove the empty sections or add content to the section.
##[error]PowerShell exited with code '1'.

### Pipeline: https://dev.azure.com/azure-sdk/public/_build/results?buildId=6707528

Failing checks on the pull request
  python - pullrequest [FAILURE] https://dev.azure.com/azure-sdk/29ec6040-b234-4e31-b139-33dc4287b756/_build/results?buildId=6707528
  python - pullrequest (Build Analyze) [FAILURE] https://dev.azure.com/azure-sdk/29ec6040-b234-4e31-b139-33dc4287b756/_build/results?buildId=6707528&view=logs&jobId=b70e5e73-bbb6-5567-0939-8415943fadb9

Copilot detected the failing pipeline and generated the analysis above. To have it attempt a fix automatically, reply with @copilot please fix the failing pipeline on this PR.

Generated by Pipeline Analysis - Next Steps · 19.1 AIC · ⌖ 6.2 AIC · ⊞ 6.6K · ◷

[Hashim Khan (Hashim1999164)]

Analyze failed because the validator treated 2.1.0b1 as the changelog being released. That dated entry was older than the unreleased 2.1.0b2 heading, and an earlier revision left an empty Bugs Fixed section under 2.1.0b1.

That is cleaned up. 2.1.0b1 (2026-08-11) stays as the released version with no empty sections. The history limit fix is under 2.1.0b2 (Unreleased), and the package version is now 2.1.0b2 so Analyze validates that unreleased entry.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated no new comments.

Suppressed comments (2)

sdk/agentserver/azure-ai-agentserver-responses/azure/ai/agentserver/responses/store/_memory.py:329

• The new inline comment says “history items”, but resolved is a mix of history_item_ids, input_item_ids, and output_item_ids (per the method’s own logic). To avoid misleading future readers, update the comment (and ideally the docstring) to reflect that this keeps the most recent N item IDs from the resolved chain, while preserving chronological order within the returned slice.

            # Keep the most recent N history items (drop oldest when over limit).
            return resolved[-limit:]

sdk/agentserver/azure-ai-agentserver-responses/azure/ai/agentserver/responses/store/_file.py:583

• Same as _memory.py: this comment refers to “history items”, but the function returns IDs that can include history, input, and output item IDs. Update wording (and any nearby docstring text if present in this file) to match the actual returned content and the “keep newest” semantics.

            # Keep the most recent N history items (drop oldest when over limit).
            return resolved[-limit:]

[Hashim Khan (Hashim1999164)]

Updated the comments and nearby docs. They now say this keeps the most recent N item IDs from the resolved chain (history, input, and output), preserving chronological order in the returned slice.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated no new comments.

[Shiva S (Shivakishore14)]

Thanks for the PR Hashim Khan (@Hashim1999164) , much appreciated