CCAR-P Advanced Agentic Architecture Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✓

  Map each section to a structured node with its clause identifiers and cross-references, summarize nodes, then synthesize using the reference graph.

  Why this is correct

  Building an explicit reference graph lets the synthesis step follow links between distant sections, so a summary of one clause can resolve references to another. Structured nodes preserve clause identifiers as anchors, and the graph makes cross-references navigable rather than lost. This directly targets the requirement by carrying relational structure through summarization instead of discarding it at chunk boundaries.

• ✗

  Increase max_tokens on each summarization call so the model can produce longer per-chunk summaries.

  Why it's wrong here

  max_tokens bounds output length, not input context, so it cannot make an oversized contract fit. Longer chunk summaries also do not restore cross-references that were severed by splitting. The bottleneck is the input window and the loss of relational structure, neither of which a larger output budget addresses, making this a fix for the wrong constraint.

• ✗

  Use a smaller model with a longer context window to process the whole contract in a single pass.

  Why it's wrong here

  Even a long-context model can degrade on very large inputs, and the scenario already states the document exceeds the window, so a single pass is not guaranteed to fit or to attend reliably to distant clauses. Model size is a proxy for capability, not a solution to structure loss. The requirement is preserving cross-references, which needs explicit linkage, not just more tokens.

• ✗

  Split the contract into fixed-size chunks, summarize each independently, and concatenate the summaries in order.

  Why it's wrong here

  Independent chunk summaries cannot see clauses referenced in other chunks, so cross-references like 'subject to Section 12.4' become dangling and the concatenated result reads as disconnected fragments. Fixed splitting also cuts across logical boundaries, separating a clause from its definitions. The described requirement is exactly cross-reference integrity, which independent summarization structurally cannot provide.