test: add Session/Memory/Summary replay consistency harness

[Thoth-hyx]

Summary

This PR adds a reusable cross-backend replay consistency harness for Session, Memory, and Session Summary behavior.

The harness executes the same standardized operation traces against InMemory and SQLite-backed services, normalizes non-business fields, compares the replay results, and generates a field-level JSON diff report.

Closes #

Changes

• Add 11 replay cases covering:
  • Single-turn and multi-turn conversations
  • Function calls and responses
  • Session, user, app, and temporary state updates
  • Memory storage and retrieval
  • Multi-session memory
  • Summary generation and update
  • Summary persistence and event truncation
  • Summary ownership isolation
  • Failed/retried and duplicate writes
• Add deterministic mock-based summary generation.
• Normalize generated IDs, timestamps, empty parts, collection representations, and dictionary ordering.
• Compare summary text separately from metadata such as version, update time, compression counts, and session ownership.
• Require backend-scoped and path-specific allowed_diff entries with explanations.
• Generate session_memory_summary_diff_report.json with:
  • Session ID
  • Event index or summary ID
  • Field path
  • Values from both backends
• Support:
  • REPLAY_MODE=inmemory for InMemory-only runs
  • Default InMemory + SQLite lightweight mode
  • Environment-controlled integration mode
• Add 10 mutation tests for intentionally injected inconsistencies.

No production code under trpc_agent_sdk/sessions is changed.

Validation

• Replay consistency tests: 24/24 passed
• Normal replay cases: 11/11 passed
• Injected inconsistency detection: 10/10
• Summary loss, overwrite, and wrong-session detection: 100%
• Normal-case false-positive rate: 0%
• Lightweight replay runtime: approximately 0.26 seconds
• Full tests/sessions regression suite: 361/361 passed
• Flake8: passed

How to run

# Default lightweight mode: InMemory + SQLite
pytest tests/sessions/test_replay_consistency.py -q

# InMemory-only mode
REPLAY_MODE=inmemory pytest tests/sessions/test_replay_consistency.py -q

# Integration mode
REPLAY_MODE=integration \
REPLAY_SQL_URL=<sql-url> \
REPLAY_REDIS_URL=<redis-url> \
pytest tests/sessions/test_replay_consistency.py -q

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (main@8080800). Learn more about missing BASE report.

Additional details and impacted files

@@            Coverage Diff             @@
##             main        #216   +/-   ##
==========================================
  Coverage        ?   87.92015%           
==========================================
  Files           ?         479           
  Lines           ?       44984           
  Branches        ?           0           
==========================================
  Hits            ?       39550           
  Misses          ?        5434           
  Partials        ?           0           

☔ 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.

[github-actions]

CLA Assistant Lite bot All contributors have signed the CLA ✍️ ✅

[Thoth-hyx]

I have read the CLA Document and I hereby sign the CLA