feat: add tags support to tools for A2A AgentCard skill metadata

[anuyog1004]

Description

Adds support for tags to tools, enabling them to be populated in the A2A AgentCard skill metadata.

Related Issues

Resolves #1261

Documentation PR

N/A. This is an additive feature to the existing @tool decorator. Usage is demonstrated in the PR description.

Type of Change

Bug fix

Testing

How have you tested the change?

• All existing tests pass (1877 tests)
• Added 4 new tests covering:
  • Tool decorator with tags
  • Tool decorator without tags (backward compatibility)
  • Tool decorator with multiple tags
  • A2A server tags propagation
• Ran hatch fmt --formatter and hatch fmt --linter - all checks passed
• Ran hatch test - all tests passed
• I ran hatch run prepare

Checklist

• I have read the CONTRIBUTING document
• I have added any necessary tests that prove my fix is effective or my feature works
• I have updated the documentation accordingly
• I have added an appropriate example to the documentation to outline the feature, or no new docs are needed
• My changes generate no new warnings
• Any dependent changes have been merged and published

---

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[anuyog1004]

The integration test failures are unrelated to this PR:

• 3 tests timed out (>45s) due to external API latency
• 215 integration tests passed
• All unit tests passed locally (1877 tests)

These specific failing tests:

• test_rate_limit_throttling_integration_no_retries - Tests OpenAI rate limiting behavior
• test_swarm_resume_from_executing_state - Tests multi-agent swarm execution
• test_swarm_execution_with_string - Tests swarm with string inputs

The changes in this PR only add a tags field to tools and don't modify any integration test behavior.

[anuyog1004]

Hi,

Can I please get a review on this. Thanks!

[pgrayy]

Nit: This test is already covered by test_tool_decorator_with_tags

[anuyog1004]

Addressed

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[pgrayy]

Thank you for your contribution and sorry for the delay. Overall, PR looks good, but would you be able to run hatch run prepare to address the lint failures. Also there is a nit called out in the unit tests.

[anuyog1004]

Hi @pgrayy, thanks for reviewing. I have addressed the nit comment on the unit test.
Regarding the other comment, hatch run prepare is working correctly for me locally. Here's the output :

-- Docs:
  https://docs.pytest.org/en/stable/how-to/capture-warnings.html
  -------------- generated xml file:
  /Users/anuyogrohilla/strands/strands-sdk-python/build/test-results.xml ---------------
  ==================================== 1872 passed, 4 skipped, 206 warnings in 12.29s
  =====================================

[pgrayy]

Sorry I didn't catch this in the first review, but I'm not so sure the tool spec is a safe spot for the tags field. Model providers might throw an error if provided. I think storing the tags on the tool object itself would be better.