frostbyte_neo
0 关注者 · 0 关注中
  • 加入于 2026-07-22
仓库列表
1
 项目
1
 软件包 代码 公开活动 已点赞
2026-09-05 16:41:29 +00:00
079ead6319 chore: append 2026-09-05 skill-collections changelog entry
89a53518d8 chore: update SKILL COLLECTIONS table — Sep 05 2026 run
比较 2 提交 »
2026-09-05 16:41:27 +00:00
2026-09-05 16:41:27 +00:00
ca68cbf9fe docs(subagents): update Last Updated badge to v2.1.261 (Sep 05, 2026)
d54148b321 changelog(subagents): append 2026-09-05 drift entry for v2.1.261
67f7d595d2 append v2.1.261 drift-check entry to commands changelog
70481a1ae0 add /skill-doctor to Extensions tag, update /diff description, bump badge to v2.1.261
51ef2c8278 docs(claude-skills): update Last Updated badge to Sep 05 2026 and version to v2.1.261
比较 11 提交 »
2026-09-05 16:41:25 +00:00
2026-09-05 16:22:04 +00:00
da840b6216 Add [Unreleased] section for next cycle
d981de1229 Release v0.85.1
fbc87d25a8 feat(ai): refresh generated image model catalog
cce6553987 docs(tui): document Alt-wheel scrolling in unreleased changelogs
082f7577c2 fix(coding-agent): make footer debounce tests deterministic
比较 9 提交 »
2026-09-05 16:22:04 +00:00
2ef483511d feat(tui): add native clipboard support
da840b6216 Add [Unreleased] section for next cycle
d981de1229 Release v0.85.1
fbc87d25a8 feat(ai): refresh generated image model catalog
cce6553987 docs(tui): document Alt-wheel scrolling in unreleased changelogs
比较 14 提交 »
2026-09-05 16:22:03 +00:00
2026-09-05 16:20:58 +00:00
937e59a547 release: 0.9.54 — dup-edge collapse, unique memory-doc names, top-level hyperedge export, rationale scoring tier, atomic version/canvas writes, tomli runtime dep
782124c7a1 chore(lock): sync uv.lock for the tomli runtime dependency (#3306)
a17a4bb8ab chore(readme): update Discord invite link
7841244f44 docs(manifest): describe tomli-missing path as fail-visible, not fail-loud
d424a7ccfa fix(manifest): require tomli on Py3.10 and fail loud when missing
比较 17 提交 »
2026-09-05 16:20:36 +00:00
b9bf63c137 Merge pull request #1708 from yvgude/fix/3.10.1-release-copy
4fc63992a3 Merge PR #1713 into the 3.10.1 release candidate
c78bdd57c1 Merge pull request #1713 from JulienJBO/mes-2189-eviter-le-scan-global-des-sessions-lean-ctx-a-chaque
cb86563078 Merge github/main into fix/3.10.1-release-copy
fe92bb5d49 Merge pull request #1712 from yvgude/ci/lint-python
比较 26 提交 »
2026-09-05 15:59:43 +00:00
0237190978 fix: [missing-model] ofox: openai/gpt-6-astra
2026-09-05 15:47:56 +00:00
2026-09-05 15:47:53 +00:00
1d6edd8c3a chore(sync): update Tinfoil model catalog
2026-09-05 15:47:48 +00:00
2026-09-05 15:47:46 +00:00
754bc4b624 chore(sync): update LLM Gateway model catalog (#6372)
2557a4de0a chore(sync): update DevPass (LLM Gateway) model catalog (#6375)
435b72a4dc chore(sync): update NanoGPT model catalog (#6374)
48e596efb1 chore(sync): update OpenRouter model catalog (#6371)
a2b42d1f12 chore(sync): update Eden AI model catalog (#6373)
比较 99 提交 »
2026-09-05 15:47:35 +00:00
d493c15992 chore(sync): update Inceptron model catalog
2ff01a6583 chore(sync): update OpenRouter model catalog (#6368)
c51e62240e chore(sync): update Merge Gateway model catalog (#6367)
5c600a0374 chore(sync): update OpenRouter model catalog (#6366)
0f8b63b030 chore(sync): update NanoGPT model catalog (#6364)
比较 95 提交 »
2026-09-05 15:46:55 +00:00
993f503d79 docs: explain derived context budget and retired override
76f31c9c5d test: drop redundant retired config field assertions
7102347e0a refactor(agent): remove context block budget override
比较 3 提交 »
2026-09-05 15:41:06 +00:00
2026-09-05 15:41:00 +00:00
c08172a641 fix(cli): skip WebUI bundle check in dev mode
63cc5f270d refactor: remove unused internal helpers and state (#5667)
f37ba60ed4 refactor(agent): remove context block budget override (#5668)
5229989f28 feat: make context compaction visible in channels (#5656)
比较 4 提交 »
2026-09-05 15:40:52 +00:00
0e14437586 fix: nightly test failing (fla bump, kernel nightly install, peft api change) (#3980)
c38195711c fix(qwen3_5): reshape position_ids in get_cu_seqlens so sample packing works and misc staleness fixes (#3975) [skip ci]
57762262db fix(fsdp2): train every LoRA shard under cpu_ram_efficient_loading; init dist before load for EP (#3977) [skip ci]
c58f4522fc Add configurable FP8 scaling recipes (#3969) [skip ci]
0d2127a06f fix(attention): don't auto-upgrade to FA4 under context parallelism (#3928) [skip ci]
比较 5 提交 »
2026-09-05 15:18:16 +00:00