CCAR-P Advanced Agentic Architecture Practice Question
Your agent orchestrates a research task by spawning several subagents, each with its own Claude conversation, and merging their outputs. You observe that the final synthesis contradicts the subagents' findings. Which architectural change most reliably preserves fidelity when merging?
⚠ Common exam trap
The trap here is treating synthesis as a creative blending step, when the contradiction stems from lost provenance and is fixed by passing attributed evidence rather than by re-running or loosening generation.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Pass each subagent's structured findings, including source citations and confidence, into the synthesizer as explicit quoted evidence.
Fidelity through a merge depends on provenance. When the synthesizer receives each subagent's findings as attributable, cited evidence, it can align claims to sources, surface real disagreements, and avoid fabricating a consensus. Structured handoff preserves the trail from subagent output to final answer, which is precisely what a paraphrased or mutable handoff destroys.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Pass each subagent's structured findings, including source citations and confidence, into the synthesizer as explicit quoted evidence.
Why this is correct
When the synthesizer receives the subagents' outputs as attributed evidence rather than paraphrase, it can trace each claim to its origin and detect genuine conflicts instead of inventing a blended narrative. Structured findings with citations preserve provenance through the merge, letting the synthesizer reconcile or flag disagreements explicitly rather than silently overwriting them, which is what caused the contradiction.
- ✗
Raise the synthesizer subagent's temperature so it can creatively reconcile conflicting viewpoints.
Why it's wrong here
Creativity is the opposite of fidelity here; higher temperature encourages the synthesizer to smooth over or invent reconciliations, worsening the contradiction. The failure is loss of provenance, not lack of imaginative blending. Reconciliation should be evidence-driven and conservative, so increasing sampling randomness moves the system further from the desired faithful merge.
- ✗
Have the synthesizer re-run each subagent's task itself to verify the results before merging.
Why it's wrong here
Re-running every subagent duplicates cost and latency and introduces new nondeterminism, so the synthesizer may reach yet another conclusion rather than reconciling the originals. The goal is faithful aggregation of existing findings, not independent re-derivation. This also breaks the parallelism benefit that motivated subagents, and divergent re-runs can deepen rather than resolve the contradiction.
- ✗
Give all subagents access to a shared mutable scratchpad so they can overwrite each other's conclusions during execution.
Why it's wrong here
A shared mutable scratchpad introduces race conditions and lets one subagent clobber another's findings before the merge, destroying the very evidence the synthesizer needs. Concurrency without ownership guarantees inconsistent state. The problem is preserving each subagent's output through synthesis, so a mutable shared surface makes provenance harder to maintain, not easier.
About these practice questions
This CCAR-P question is part of Courseiva's 262-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Anthropic exam blueprint
This CCAR-P practice question is part of Courseiva's free Anthropic certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the CCAR-P exam.