Kiloloop Research https://kiloloop.com/research/ Research from operating a multi-runtime agent fleet: runtime capability comparisons and operational reviews of the fleet's own audit records. en-us Mon, 07 Sep 2026 12:00:00 GMT Estimate accuracy, measured: seven months of dispatch estimates against their actuals https://kiloloop.com/research/estimate-accuracy-measured/ https://kiloloop.com/research/estimate-accuracy-measured/ Mon, 07 Sep 2026 12:00:00 GMT Ratio distribution, period and month trends, declared envelope vs expected wall, size and task-class breakdowns, and verdicts from 606 dispatch rows. Reproducible with estimate_stats.py. estimation ops review agent fleet Human in the loop, measured: four months of autonomy-gate records https://kiloloop.com/research/human-in-the-loop-measured/ https://kiloloop.com/research/human-in-the-loop-measured/ Sat, 05 Sep 2026 12:00:00 GMT Pause rates, human outcomes and latency, checkpoint breaches and declared-vs-actual fidelity from 602 audit records. Reproducible with hil_stats.py. autonomy ops review agent fleet What each runtime can actually do https://kiloloop.com/research/runtime-comparison/ https://kiloloop.com/research/runtime-comparison/ Thu, 27 Aug 2026 12:00:00 GMT Claude Code, Codex, and ZCode on 17 capabilities, in each runtime's own words — plus where each is strongest and the five parity gaps that block the loop. runtime comparison agent fleet