feat(testing): add declarative scenario runner for agent evaluation

[diiviikk5]

Summary

closes #1279

Add a declarative ScenarioRunner to mofa-testing that loads YAML/JSON scenario specs, configures the existing mocks automatically, runs an async scenario body, and validates structured expectations into a TestReport.

Builds On my previous prs

• feat(testing): Implement Agent testing framework #486 - core mofa-testing mocks and framework
• feat(testing): add failure injection, sequenced responses, rate limiting & MockClock to mofa-testing #888 - failure injection, response sequencing, and MockClock
• feat(testing): add test report generator with builder, formatters #895 - test report generation and formatting

What This Adds

• ScenarioSpec::from_yaml_str and ScenarioSpec::from_json_str
• a new scenario module in mofa-testing
• ScenarioContext for configured mock backend, bus, clock, and tool access
• expectation evaluation for:
  • total infer calls
  • prompt substring counts
  • tool call counts
  • bus messages by sender
• end-to-end tests for parsing, successful execution, expectation failures, and injected execution failures

Why This Matters

mofa-testing already provides strong low-level primitives. This PR adds a higher-level declarative layer that makes end-to-end agent evaluation reproducible and easier to adopt across contributors. It is a direct step toward a real testing and evaluation platform for MoFA agents.

How I Tested

• cargo test -p mofa-testing --test scenario_runner_tests
• cargo test -p mofa-testing --test integration

[diiviikk5]

@lijingrs @yangrudan builds towards the idea 6 / testing framework also towards the agent testing framework in the open task and 3 of my previous prs , starting from #486 , #888 , #895 , furthermore #1288 extends this

[diiviikk5]

Can someone review/ merge this , i am building on top of these and #1441 and 3-4 other open prs since weeks
so to progress further if i could know the changes in these or complete merge it , would be helpfull @lijingrs