Fix/misalignment qa housekeeping

[rjavadi]

Summary

After merging Misalignment QA use case there were a few things that needed refactoring just to remain consistent with the rest of the codebase. No significant changes.

Clickup Ticket(s): Link(s) if applicable.

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)
• ✨ New feature (non-breaking change that adds functionality)
• 💥 Breaking change (fix or feature that would cause existing functionality to not work as expected)
• 📝 Documentation update
• 🔧 Refactoring (no functional changes)
• ⚡ Performance improvement
• 🧪 Test improvements
• 🔒 Security fix

Changes Made

• Move misalignment QA result exploration helpers (analysis.py) from the library package to implementations/misalignment_qa/, where they are used
• Update 02_inspect_results.ipynb to import from the new location
• Tighten MisalignmentTask agent parameter type from Any to LlmAgent
• Move load_dotenv() from run_experiment_config() to the CLI entrypoint (evaluate.py)

Testing

• Tests pass locally (uv run pytest tests/)
• Type checking passes (uv run mypy <src_dir>)
• Linting passes (uv run ruff check src_dir/)
• Manual testing performed (describe below)

Manual testing details:

• Run uv run python implementations/misalignment_qa/evaluate.py --config implementations/misalignment_qa/configs/bootcamp_misalignment.yaml --variant-id <one-variant> from repo root
• Open and run setup cells in 02_inspect_results.ipynb; confirm MisalignmentResultsExplorer imports correctly
• Verify env vars (e.g. GOOGLE_API_KEY, Langfuse keys) are still picked up when running via CLI

Checklist

• Code follows the project's style guidelines
• Self-review of code completed
• Documentation updated (if applicable)
• No sensitive information (API keys, credentials) exposed