v0.4.1

Core feature highlights

• External environment imports: added openenv import support for ORS/OpenReward and Verifiers environments, including extracted wrapper templates, safer vendoring behavior, task API integration, and an imported-environment example.
• QED Math environment: added a new QED Math env with MCP tools, LLM-judge grading, math verifier service support, reward/rubric metrics, docs, and an inference example.
• Current client/runtime docs: modernized tutorial notebooks and docs around the async EnvClient, WebSocket transport, .sync() wrapper usage, and stale Echo imports.
• Release polish and compatibility: fixed openenv serve exit semantics, refreshed FinQA's minimum OpenEnv version, refactored inspect eval kwargs, and unpinned Hugging Face Hub in HF sandbox examples now that SandboxPool ships in 1.22.0.

What's Changed

• Bump 0.4.1.dev0 by @burtenshaw in #910
• [codex] fix openenv serve exit semantics by @burtenshaw in #907
• [codex] bump finqa env minimum openenv version by @burtenshaw in #909
• [codex] fix stale echo docs imports by @burtenshaw in #908
• [codex] refactor inspect eval kwargs by @burtenshaw in #912
• Unpin huggingface_hub in HF sandbox examples by @sergiopaniego in #913
• Modernize tutorial for async EnvClient API by @vedthebear in #897
• feat: QED Math Environment with MCP tools and LLM-judge rubric by @rycerzes in #865
• [feature] Import external envs (ors, verifiers) by @burtenshaw in #726
• Release 0.4.1 by @burtenshaw in #918

New Contributors

• @vedthebear made their first contribution in #897

Full Changelog: v0.4.0...v0.4.1