Add learnings from code review analysis (#2280)

Co-authored-by: openhands <openhands@all-hands.dev>

Author: neubig

AGENTS.md

@@ -337,6 +337,7 @@ These are enforced by `check_sdk_api_breakage.py` (runs on release PRs). Depreca
 - DON'T write TEST CLASSES unless absolutely necessary!
 - If you find yourself duplicating logics in preparing mocks, loading data etc, these logic should be fixtures in conftest.py!
 - Please test only the logic implemented in the current codebase. Do not test functionality (e.g., BaseModel.model_dumps()) that is not implemented in this repository.
+- For changes to prompt templates, tool descriptions, or agent decision logic, add the `integration-test` label to trigger integration tests and verify no unexpected impact on benchmark performance.
 
 # Behavior Tests
 
@@ -423,4 +424,5 @@ For examples that use the critic model (e.g., `34_critic_example.py`), the criti
 - Ruff ignores `ARG` (unused arguments) under `tests/**/*.py` to allow pytest fixtures.
 - Repository guidance lives in `AGENTS.md` (loaded as a third-party skill file).
 </REPO_CONFIG_NOTES>
+
 </REPO>