https://handuo.top/tags/evaluation/EvaluationHugoBlox Kit (https://hugoblox.com)en-usTue, 31 Mar 2026 00:00:00 +0000https://handuo.top/media/icon_hu_39ba2e8de423d957.pnghttps://handuo.top/tags/evaluation/https://handuo.top/project/prax-agentos/Tue, 31 Mar 2026 00:00:00 +0000https://handuo.top/project/prax-agentos/<p>I joined Apodex in late March 2026 as one of three founding engineers of Prax / AgentOS and became the technical owner of ApodexHarness infrastructure. The project develops an execution system for complex, long-running tasks.</p> <p>I built <strong>heavy_mode</strong>, which combines parallel sub-agents, cross-branch synthesis, global verification, and continuation after interruption. I proposed planner-generated, schema-constrained workflow DAGs and replaced a LangGraph dependency with a compact execution engine supporting branches and verification-retry loops.</p> <p>I also developed token-budgeted context compression, file-based research memory, tool-specific result budgets, and evidence checks for unsupported numbers and citations. Reliability work included durable agent messages, sandbox isolation, tool permissions, and cancellation-safe event streaming.</p> <p>The harness connects workflows to a unified evaluation interface. I supported the Apodex 1.1 and 1.2 model releases and ran 240 APEX Agents 1.1 tasks end to end to validate the evaluation path.</p>