PROJECT TWO - ADDING CONTEXT MANAGEMENT

ARTIFICIAL INTELLIGENCE ASSISTANT PROMPTS Replace each bracketed placeholder with your own material before use. [START OF PROMPT] Review this Resume Review Agent lab's context_builder.py and compaction trigger for correctness, specifically checking whether any resume-to-posting match decision or progress-notes fact could be silently lost during compaction: [PASTE LAB context_builder.py / compaction CODE HERE]. Output format: a correctness review (Risk, Evidence, Fix) plus a confirmation of whether progress_notes.md would correctly allow resume-after-restart given this code. [END OF PROMPT] [START OF PROMPT] Design a before/after measurement experiment for our Stage 2 context-management addition, using our 25-resume sample batch, and then interpret these actual results: fixed task set = lab's 25 synthetic resumes, before/after data = [PASTE TOKEN COUNT, COST, AND SUCCESS-RATE NUMBERS HERE]. Output format: an experiment design summary plus an interpretation stating whether the results show a real, meaningful improvement or are within normal noise. [END OF PROMPT]