Remove dead chat code and obsolete planning docs.

feat: robust marker-name matching + cleaned real training labels

feat(eval): add real lab-report eval data and KB references

feat(eval): field-level extraction eval harness + tests