| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| reef_infra-0.1.2-py3-none-any.whl | 2026-10-02 | 1.4 MB | |
| reef_infra-0.1.2.tar.gz | 2026-10-02 | 8.6 MB | |
| README.md | 2026-10-02 | 7.0 kB | |
| v0.1.2 source code.tar.gz | 2026-10-02 | 8.6 MB | |
| v0.1.2 source code.zip | 2026-10-02 | 9.7 MB | |
| Totals: 5 Items | 28.3 MB | 0 | |
What's Changed
- docs(agents): require the coding style checklist before submitting a PR by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/655
- Update header for clarity in README.md by @quao627 in https://github.com/Human-Agent-Society/reef/pull/654
- ci(review): replace the Merge Oncall with CODEOWNERS review routing by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/657
- fix(harness): the hermes adapter delivers the model key through the home's .env file by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/595
- fix(codex): propagate served model metadata to harnesses by @jane-jhu in https://github.com/Human-Agent-Society/reef/pull/658
- fix(artifacts): a fork pushes over history whose LFS objects it never fetched by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/644
- fix(docs): update broken SGLang installation guide link by @kuishou68 in https://github.com/Human-Agent-Society/reef/pull/659
- fix(harness): keep reef-claude sessions off the person's own Claude Code install by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/633
- docs: require documentation review before pull requests by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/660
- fix(connector): report each scenario's adapter, step summaries and harness requests by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/635
- chore(deps-dev): bump the python-minor-and-patch group with 2 updates by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/669
- chore(deps-dev): update connectrpc requirement from <0.12,>=0.11.1 to >=0.11.1,<0.13 by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/670
- fix(records): name the release a pulling client runs on its inference records by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/640
- fix(artifacts): a release saved for review pushes over history whose LFS objects it never fetched by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/661
- feat(sdpo): the SDPO recipe on the shared distillation base with SciKnowEval Chemistry as its example by @MikeZheng777 in https://github.com/Human-Agent-Society/reef/pull/649
- chore(deps): bump the docs-site-minor-and-patch group in /docs/site with 5 updates by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/671
- Complete Guidance-TTT examples and document four discovery tasks by @Chonghe-Jiang in https://github.com/Human-Agent-Society/reef/pull/647
- fix: correct harness reinstalls, token setup, and connection messages by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/641
- fix(harness): check the installed tree before a session, keep session copies out of TMPDIR, and never follow a link in the tree by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/639
- docs: keep documentation focused on reader-facing interfaces by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/684
- feat(train): score centering as a correction term on policy-gradient loss families (RFC [#672]) by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/673
- fix(docker): upgrade FlashInfer to 0.7.0 so SGLang's attention matches the trained policy by @MikeZheng777 in https://github.com/Human-Agent-Society/reef/pull/685
- fix(opencode): grade episodes from opencode.db, add web search to sessions, and keep model calls on Reef by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/638
- fix(harness): codex skills and commands in sessions, approvals for calls to Reef, and graded replies by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/634
- fix(record2dataset): the reply check accepts file names that start with an underscore by @simonucl in https://github.com/Human-Agent-Society/reef/pull/686
- fix(tests): keep the docs terminology check passing on main by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/687
- docs: guide concise issue and RFC writing by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/690
- fix(dsh): boot reef-dsh web, keep the tree's commands user only, and proxy the anthropic binding by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/636
- fix(terminus): route every served model through litellm_proxy and run reefine evaluation on Docker or E2B by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/631
- fix(harness): keep every model call on Reef's model binding by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/632
- fix(harness): reef-hermes sessions load release commands, keep their state, identity and resume command, and turn off tirith, the skill review and the curator by @Benjamin-eecs in https://github.com/Human-Agent-Society/reef/pull/637
- feat(train): support score centering for sampled OPD by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/689
- chore(deps): bump pyjwt from 2.13.0 to 2.14.0 by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/692
- feat(harness): support E2B episode execution by @BobbyZhouZijian in https://github.com/Human-Agent-Society/reef/pull/605
- chore(deps): bump brace-expansion in /docs/site by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/693
- fix(harness): wait for steps triggered by passing reports by @GBX-Max1220 in https://github.com/Human-Agent-Society/reef/pull/677
- feat(inference): add vLLM engine launch and control behind the inference service contracts by @ruizhang0101 in https://github.com/Human-Agent-Society/reef/pull/646
- fix: scenario deleted during harness step ends without commit error (#628) by @viwe-monai in https://github.com/Human-Agent-Society/reef/pull/696
- fix(train): say how far a partly filled batch still has to go by @MikeZheng777 in https://github.com/Human-Agent-Society/reef/pull/697
- chore(deps): bump dompurify from 3.4.14 to 3.4.16 in /docs/site by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/703
- chore(deps): bump virtualenv from 21.7.5 to 21.7.12 by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/704
- chore(deps): bump pyjwt from 2.14.0 to 2.15.0 by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/705
- chore(deps): bump urllib3 from 2.7.0 to 2.8.0 by @dependabot[bot] in https://github.com/Human-Agent-Society/reef/pull/706
- docs(harness): define rendering and regroup the harness tree node kinds by @yanxz in https://github.com/Human-Agent-Society/reef/pull/708
New Contributors
- @kuishou68 made their first contribution in https://github.com/Human-Agent-Society/reef/pull/659
- @GBX-Max1220 made their first contribution in https://github.com/Human-Agent-Society/reef/pull/677
- @viwe-monai made their first contribution in https://github.com/Human-Agent-Society/reef/pull/696
Full Changelog: https://github.com/Human-Agent-Society/reef/compare/v0.1.1...v0.1.2
