Generalize settings and logs assistant

AGENTS.md

@@ -57,7 +57,7 @@ A spec/change is not considered done merely because code is merged, tests pass,
 
 Before archiving any OpenSpec change, perform a refactoring pass over the code and specs touched by that change. The pass must inspect both the current diff and the surrounding implementation context, because a small diff may reveal repeated patterns or structural problems that only become obvious when compared with nearby code.
 
-Do the refactoring pass in these distinct areas, first start a subagent for each area to identify potential improvements in parallel, then implement them:
+Do the refactoring pass in these distinct areas in sequence so they are less likely to converge on the same issues, first start a subagent to identify potential improvements using `gpt-5.3-codex` if available, then implement them, then start the repeat for the next area:
 
 1. Check for DRYness. Look for duplication introduced by the change and for existing nearby duplication that the change now makes worth consolidating. A change may be small on its own, but if it is the fifth copy of the same idea, it is a refactoring target.
 2. Check for SOLID violations. Look for responsibilities that are mixed together, abstractions that are hard to replace or test, interface shapes that force unrelated dependencies, and code paths that require modifying stable code for each new variant.
@@ -85,6 +85,10 @@ If production or end-to-end verification is not possible, state exactly why and
 
 ## Implementation Notes
 
+Meeting Assistant is Manuel's local meeting capture and knowledge system. It runs as a background .NET service with a Windows tray icon, captures microphone plus system audio, transcribes meetings, maintains Obsidian meeting artifacts, enriches meetings from Outlook metadata when available, applies local workflow rules, identifies speakers, captures screenshots, and runs agentic summarization against a LiteLLM/OpenAI-compatible endpoint.
+
+The app is used during real meetings. Treat active recording, transcription finalization, OCR, and summarization as live user work. Prefer local endpoints, generated note files, and application logs for verification, and avoid restart or cleanup actions that could interrupt an active run unless the user explicitly requested them.
+
 Keep source and deployment ownership separate:
 
 - `Manuel/meeting-assistant` owns application source, tests, build configuration, container image publishing, and OpenSpec source changes.