fix(integrations): add values for pydantic-ai and openai-agents to _INTEGRATION_DEACTIVATES to prohibit double span creation

[constantinius]

Issues

Contributes to: https://linear.app/getsentry/issue/TET-1511/openai-agents-double-span-reporting

[linear]

TET-1511 OpenAI Agents: double span reporting

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.20%. Comparing base (9a1b1aa) to head (5178a11).
⚠️ Report is 1 commits behind head on master.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #5196      +/-   ##
==========================================
+ Coverage   82.94%   84.20%   +1.25%     
==========================================
  Files         181      181              
  Lines       18441    18441              
  Branches     3283     3283              
==========================================
+ Hits        15296    15528     +232     
+ Misses       2130     1899     -231     
+ Partials     1015     1014       -1     

Files with missing lines	Coverage Δ
sentry_sdk/integrations/__init__.py	86.95% <ø> (-1.09%)	⬇️

... and 12 files with indirect coverage changes

[cursor]

Bug: Test doesn't verify OpenAI/Anthropic integrations are enabled

The test test_disabling_integrations_allows_openai_and_anthropic was expanded to disable OpenAIAgentsIntegration and PydanticAIIntegration in addition to LangchainIntegration, but the assertion still only verifies that LangchainIntegration not in integration_types. The test name implies it verifies that OpenAIIntegration and AnthropicIntegration are allowed to be enabled, but there's no assertion checking that they actually are in integration_types.