Instrument requirement validators to track validation attempts and failures

[psschwei]

Description:
Instrument requirement validators to track validation attempts and failures.

Detailed Requirements:

1. Add spans for validation:
   • validator.check - Individual validation check
2. Add attributes:
   • requirement_type - Type of requirement (python, markdown, llmaj, etc.)
   • requirement_name - Specific requirement identifier
   • status - pass/fail
   • failure_reason - Reason for failure (on fail)
3. Add span events for:
   • Validation start with input summary
   • Validation result with details
4. Instrument all requirement types:
   • PythonRequirement
   • MarkdownRequirement
   • ToolCallRequirement
   • ALoraRequirement
   • LLMaJRequirement

Files to Modify:

• mellea/stdlib/requirements/base.py - Base instrumentation
• mellea/stdlib/requirements/python.py - Python requirement
• mellea/stdlib/requirements/markdown.py - Markdown requirement
• mellea/stdlib/requirements/tool_call.py - Tool call requirement
• mellea/stdlib/requirements/alora.py - ALoRA requirement
• mellea/stdlib/requirements/llmaj.py - LLMaJ requirement

Acceptance Criteria:

• Each validation check has a span
• Pass/fail status captured
• Failure reasons recorded
• All requirement types instrumented

[ajbozarth]

Closing as part of the tracing epic (#444) re-plan. This sub-issue predates the plugin/hooks pattern (PR #653) and the GenerationMetadata partition (PR #908); its spec has drifted.

Replacement: part of #1050 — sampling, validator, and formatter spans via plugin hooks (rolled together since the mechanism is identical across all three).

See #444 for the full refreshed plan.