Conventional reasoning protocols fix a pre‑selected task, so they cannot test whether an agent propagates relevant updates, preserves unaffected work, or reconstructs a historical task binding. We therefore define dynamic task routing: an event stream continuously revises task bindings, and the system must select the document version valid at each query time before solving it. To evaluate this, we repurpose six popular benchmarks (MMLU, MMLU‑Pro, MedMCQA, MATH, GPQA, HumanEval) into 31,119 dynamic episodes containing 373,428 temporally categorized queries. This setting exposes a core trade‑off: recomputing after every event wastes work, while unchecked reuse yields stale conclusions.
We introduce the Revision‑Aware Independent Agent Graph (RIAG), a bounded multi‑agent policy that separates deterministic temporal resolution from task reasoning. RIAG caches solutions by immutable document identity, starts each fresh task with two unexposed attempts, and conditionally invokes audit and repair, using at most four calls per document version. On the full collection, homogeneous RIAG achieves 54.24% joint routing‑and‑answer accuracy with 0.62 calls per query, compared to 32.22% at 18.00 calls per query for the strongest baseline; heterogeneous RIAG reaches 49.78% at 0.63 calls per query.
Review