fix(supermemory): honor auto_capture on session end

[martinramos002-bot]

Summary

• Make Supermemory on_session_end() return early when auto_capture is disabled.
• Add a regression test so session-end conversation ingest follows the same capture opt-out as per-turn sync.

Test Plan

• python -m pytest -o addopts='' tests/plugins/memory/test_supermemory_provider.py -q
• python -m ruff check plugins/memory/supermemory/__init__.py tests/plugins/memory/test_supermemory_provider.py

No private data or local paths are included; the test uses synthetic fixture messages.

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Updates the Supermemory provider to skip ingesting session messages when auto-capture is disabled, and adds a regression test to confirm this behavior.

Changes:

• Gate on_session_end ingestion behind the _auto_capture flag.
• Add a unit test ensuring no ingest occurs when _auto_capture is False.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
tests/plugins/memory/test_supermemory_provider.py	Adds coverage for on_session_end behavior when auto-capture is disabled.
plugins/memory/supermemory/init.py	Updates on_session_end to respect the _auto_capture setting.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[martinramos002-bot]

Thanks — addressed the Copilot nit in the latest push (178da018):

• Updated the regression test to configure auto_capture: false through the provider config file before initialize(), rather than mutating the private _auto_capture attribute directly.
• Re-ran validation:
  • python -m pytest -o addopts='' tests/plugins/memory/test_supermemory_provider.py -q → 30 passed
  • python -m ruff check plugins/memory/supermemory/__init__.py tests/plugins/memory/test_supermemory_provider.py → passed

[alt-glitch]

Duplicate of #24136 (same one-line _auto_capture guard in on_session_end()). Both are subsumed by #21692 which includes this fix plus additional memory bounds.

[martinramos002-bot]

Thanks for the pointer — agreed. This is covered by #24136 and more broadly by #21692, so I’m closing this duplicate to avoid competing review threads. We’ll track #21692 for absorbing the local Supermemory auto_capture safeguard once it lands in an official release.