[codex] Improve high-event hot paths

[helto4real]

Breaking change

None. This keeps public APIs unchanged and preserves raw OnHassMessage publication behavior.

Proposed change

This PR implements the measured high-event performance improvements from the performance review plan.

The command/result hot path now uses an ID-indexed pending-result table instead of creating one filtered Rx subscription per command. HomeAssistantConnection registers a TaskCompletionSource<HassMessage> for each command id, completes matching result messages directly in the receive loop, and still publishes every raw HassMessage to OnHassMessage for ping, trigger, and diagnostic consumers.

Review follow-up fixed the pending-result lifecycle: pending entries are removed when a result arrives, the connection is disposed, sending fails, the caller cancels while waiting, or the result wait times out. Tests cover timeout and cancellation cleanup.

The websocket receive path reads a complete websocket message into a byte buffer, detects whether the payload is a JSON object or coalesced array, and deserializes directly to T or T[]. This removes the prior JsonElement intermediate and second typed deserialization on the ingest path while preserving chunked message and remote-close behavior.

The benchmark harness keeps the owning JsonDocument alive for the state-change benchmark lifetime, fixing the previous invalid JsonElement baseline failure.

Review follow-up also corrected ResultDispatchBenchmarks so the implemented path uses ConcurrentDictionary and awaits all completion tasks, matching the Rx baseline's completed-work shape. The corrected local in-process result-dispatch run showed the expected fixed overhead for one pending command and large wins as pending command count grows:

Pending commands	Baseline Rx dispatch	Concurrent dictionary dispatch	Result
1	250.2 ns, 880 B	485.2 ns, 1,800 B	slower for a single waiter
10	2,175.3 ns, 6,424 B	929.4 ns, 3,312 B	57% faster, 48% lower allocation
100	30,939.1 ns, 133,144 B	7,566.0 ns, 24,880 B	76% faster, 81% lower allocation
1000	1,819.0 us, 8,528,351 B	80.2 us, 262,848 B	96% faster, 97% lower allocation

Detailed local report: docs/performance/high-event-review.md.

Type of change

• Dependency upgrade
• Bugfix (non-breaking change which fixes an issue)
• New feature (which adds functionality to an existing integration)
• Breaking change (fix/feature causing existing functionality to break)
• Code quality improvements to existing code or addition of tests

Additional information

• This PR fixes or closes issue: fixes #
• This PR is related to issue:
• Link to documentation pull request:

Checklist

• The code change is tested and works locally.
• Local tests pass. Your PR cannot be merged unless tests pass
• The code compiles without warnings (code quality check)
• Tests have been added to verify that the new code works.

If user exposed functionality or configuration are added/changed:

• Documentation added/updated for https://netdaemon.xyz

No user-facing functionality or configuration was added or changed.

[codecov]

Codecov Report

❌ Patch coverage is 96.47059% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 81%. Comparing base (b913b8f) to head (73a4af3).

Files with missing lines	Patch %	Lines
...mon.HassClient/Internal/HomeAssistantConnection.cs	95%	1 Missing and 1 partial ⚠️
...sClient/Internal/Net/WebSocketTransportPipeline.cs	96%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@         Coverage Diff          @@
##           main   #1389   +/-   ##
====================================
- Coverage    81%     81%   -1%     
====================================
  Files       202     202           
  Lines      3941    3973   +32     
  Branches    444     453    +9     
====================================
+ Hits       3218    3241   +23     
- Misses      531     541   +10     
+ Partials    192     191    -1     

Flag	Coverage Δ
unittests	81% <96%> (-1%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[helto4real]

Code review summary: the optimization direction looks sound and CI is green, but the pending-result lifecycle should be fixed before merging, and the benchmark should compare equivalent production paths before the precise performance claims are retained.