Add progress indicator to oaeval audit command

[PrinceThummar011]

Summary

Adds a progress indicator to the oaeval audit command using rich.progress, consistent with the existing pattern in run.py. Users now see which corpus check is currently executing instead of an unresponsive terminal during long-running audits.

Changes

• openagent_eval/corpus/auditor.py — Introduced an optional progress_callback parameter on CorpusAuditor.audit(). The callback is invoked at three points: before document scanning begins, when each configured analyzer starts, and upon audit completion.
• openagent_eval/cli/commands/audit.py — Integrated a rich.progress.Progress context in the CLI command and connected it to the new callback to update the task description and progress bar in real time.
• tests/unit/test_cli/test_cli_improvements.py — Added TestAuditCommand, covering help text, invalid-path error handling, and verification that the progress callback fires correctly via mocking.

Design Notes

• Progress reporting is implemented as a callback rather than embedded directly in CorpusAuditor, keeping the auditor UI-agnostic and reusable outside the CLI context.
• Visual styling and column configuration mirror run.py for consistency across commands.
• Existing exit codes, error handling, and summary output are unchanged — this is a UX addition only, with no behavioral changes to audit logic.

Testing

• Full test suite passes locally.
• Note: tests/unit/test_metrics/test_generation.py::TestSemanticSimilarity failed during a full run; confirmed unrelated to this change and pre-existing on main.

Closes #123

[github-actions]

🎉 Congratulations @PrinceThummar011!

Your pull request has been successfully merged into main. 🚀

Thank you for contributing to OpenAgentHQ and helping improve the project.

We truly appreciate your contribution and hope to see you back with more amazing PRs!

Happy Open Sourcing! ❤️