What changed
Oilbird reuses verifier hidden states to retrieve semantic drafts that exact text matching misses. Across three drafters it raised accepted draft length by 24% to 29%.
Why it matters
Tool calls repeat structure while changing identifiers and values, making exact-prefix draft reuse unnecessarily brittle for agent workloads.
What to do
Track the method for self-hosted, high-volume structured tool traffic and compare end-to-end latency under your prompt lengths, concurrency, batching, and networking.
Caveat
No implementation was linked. The 4.4x API-Bank speedup is creator-measured in a research harness and may shrink under production scheduling.