// Release stream · daily archive
Wed, Sep 30
124 releases from tracked repos.
Wed, Sep 30· 82
-
anthropics/claude-code v2.1.285 Patch Desktop appNew agent
Added `CLAUDE_CODE_DISABLE_WEB_FETCH` environment variable to turn off the WebFetch tool
-
zhayujie/CowAgent 2.2.0 Minor Desktop app
🌐 English | 中文
-
code-yeongyu/oh-my-openagent v5.1.5 First release New agent
Big thanks to @ashmoonori-afk, whose #9209 teaches memory recall to find Korean, Japanese and Chinese notes and to pick the right note out of a big memory.**
-
stablyai/orca v1.4.217 Patch Desktop app
Thank you so much for using Orca and for your continued support! ❤️
-
cline/cline sdk/sdk/v0.0.88 First release New model +3 more
Direct Anthropic requests now ask the Claude API for its recommended server-side refusal fallback (`anthropic.fallbacks: "default"`), and OpenRouter and Cline requests for Anthropic models set `provider.allow_fallbacks`
-
pydantic/pydantic-ai v2.52.0 Minor New agent +1 more
This release fixes one security issue in `web_fetch`. See the advisory for full details and affected versions.
-
Comfy-Org/ComfyUI v0.38.0 Minor New model
Improve qwen 2.1 KV cache location logic by @rattus128 in https://github.com/Comfy-Org/ComfyUI/pull/16429
-
netdata/netdata v2.12.0 Minor
The v2.12.0 build of the Netdata Agent for Windows has been discovered to have serious stability issues that cause it to crash randomly on startup due to an upstream bug in some of the tooling we depend on. As such, we h
-
anywhere-labs/dsh-desktop v2.0.17-next Patch Desktop appIDE extension +2 more
本次更新升级至 DeepSeek Harness `v0.2.0-rc.2`,修复 Linux AppImage 在应用内重启后无法重新启动的问题。 Next 同时修复 Session 日志下载和 `/export` 无响应的问题。
-
Niko1221/Strata v0.1.25 First release +4 more
Faster prompts, AMD Radeon support (experimental), and a smaller KV cache option.**
-
K-Dense-AI/scientific-agent-skills v2.70.0 Minor
Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 250,000+ scientists worldwide. 177 ready-to-use validated skills plus 100+ scientific da…
-
sickn33/agentic-awesome-skills v18.10.0 Minor
[18.10.0] - 2026-09-30 - "SME Operations, 15 Official Beatra Skills, and Four New Advisories Closed"
-
kortix-ai/suna v0.13.42 Patch Desktop app +2 more
Connect your own computer from the desktop app, Teams at Slack parity, and saved history that is always on.
-
langfuse/langfuse v4.48.0 Minor
feat(seeder): add incident-session scenario for the session timeline by @nkabardin in https://github.com/langfuse/langfuse/pull/18061
-
openai/openai-python v3.22.0 Minor +2 more
3.22.0 (2026-09-29)
-
mastra-ai/mastra @mastra/[email protected] Sub-package
Live channel resolvers (no-redeploy channel connections)
-
openclaw/openclaw v2026.9.7 Rolling
OpenClaw v2026.9.7
-
Arize-ai/phoenix @arizeai/[email protected] Sub-package +1 more
AI Observability & Evaluation
-
osaurus-ai/osaurus 0.25.15 Patch Desktop app
Orchestrator audit remediation: truthful addendum, apply-first config, pool guard, default new-chat agent, compact spawn envelope (#2936) by @tpae
-
activepieces/activepieces 0.92.1 Minor
fix(engine): remove per-sandbox piece LRU (#15786) @amrdb
-
casdoor/casdoor v4.13.0 Minor +1 more
c9adca039f3c24b570567e0def8d1018e9194c9f feat: add Audit provider to forward audit logs to Syslog
-
run-llama/liteparse node-v2.15.0 Minor +3 more
fix(ocr): keep parentheses around numbers in OCR table cleanup by @SammyTourani in https://github.com/run-llama/liteparse/pull/470
-
langchain-ai/deepagents deepagents-code==0.1.79 Patch New agent +1 more
MCP tool calls now use configurable timeouts to prevent them from hanging indefinitely (#6625).
-
Significant-Gravitas/AutoGPT autogpt-platform-beta-v0.8.2 Patch New model
🔥 What's New?
-
langflow-ai/langflow v1.12.4 Patch
fix: reject mixed blocked DNS answers with IP allowlists by @erichare in https://github.com/langflow-ai/langflow/pull/15334
-
langchain-ai/langchain langchain-openai==1.6.7 Patch
Changes since langchain-openai==1.6.6
-
siyuan-note/siyuan v3.8.7-alpha.1 Patchpre Desktop app
Improve mobile shorthand submission success feedback
-
axolotl-ai-cloud/axolotl v0.20.0 Minor
Go ahead and axolotl questions
-
apify/apify-mcp-server v0.17.0 Minor +1 more
0.17.0 (2026-09-30)
-
omnigent-ai/omnigent v0.16.0 Minor
Major new features
-
google-gemini/gemini-cli v0.64.0-nightly.20260930.g38700b4b3 Minorpre +2 more
fix(core): enable autonomous plan execution in non-interactive mode by @urielefrenvirtusa in https://github.com/google-gemini/gemini-cli/pull/29539
-
ruvnet/RuView v2940 Rolling +4 more
Automated release from CI pipeline
-
google/agents-cli v1.8.0 Minor
The CLI and skills that turn any coding assistant into an expert at creating, evaluating, and deploying AI agents on Google Cloud.
-
UKGovernmentBEIS/inspect_ai 0.3.273 First release
Inspect: A framework for large language model evaluations
-
vercel/streamdown streamdown%402.7.0 Minor +5 more
A drop-in replacement for react-markdown, designed for AI-powered streaming.
-
pbakaus/impeccable engine-v0.1.8 Patch
Prebuilt impeccable engine binaries (engine-v0.1.8). The launcher, the npm shim and impeccable install download these on first run. Docs: https://impeccable.style
-
QuantumNous/new-api v1.0.0-rc.41 Patch
[!IMPORTANT]
-
herdrdev/herdr v0.9.3 Patch +1 more
This is a hotfix release for v0.9.2. See the v0.9.2 notes for the full feature release: https://github.com/herdrdev/herdr/releases/tag/v0.9.2
-
deepset-ai/haystack v3.4.0-rc0 Minorpre
Open-source AI orchestration framework for building context-engineered, production-ready LLM applications. Design modular pipelines and agent workflows with explicit control ove…
-
CopilotKit/CopilotKit v1.75.2 Patch +1 more
fix(react-core): use new run IDs for standard interrupt resumes (#7001) (081e745)
-
superlinked/sie v0.9.0 Minor
Open-source inference server and production cluster for all the models your agent needs.
-
topoteretes/cognee v1.6.2 Patch
v1.6.2 — Embedding reliability & Slack history
-
labring/FastGPT v4.17.1 Patch IDE extensionNew model
更新 fastgpt-app(FastGPT 主服务)镜像 tag: v4.17.1
-
simstudioai/sim v0.9.6 Patch
feat(library): Best ChatGPT alternatives for building AI agents and automating workflows (#8414)
-
earthtojake/text-to-cad v0.7.4 Patch
Other Changes
-
ai-dynamo/nixl v1.5.0 Minor
NIXL 1.5.0 expands transfer preparation and metadata exchange with a strided descriptor API, zero-copy index handling, and a metadata manager that adds TCPStore alongside P2P and ETCD. Python and Rust bindings expose mem
-
confident-ai/deepeval typescript-v0.9.21 Patch +1 more
The LLM Evaluation Framework
-
intel/auto-round v0.16.0 Minor
PTQ-Level Efficiency, QAT-Level Accuracy|训练后量化的效率,量化感知训练的精度
-
YishenTu/claudian 2.3.10 Patch
fix(chat): show fast mode indicator at full strength, keep reasoning muted (#1484)
-
apache/tvm v0.27.0.post1 Patch
Open Machine Learning Compiler Framework
-
elie222/inbox-zero desktop-v0.1.18 Patch
The world's best AI personal assistant for email. Open source app to help you reach inbox zero fast.
-
anthropics/claude-agent-sdk-typescript v0.3.285 Patch New agent
Added an optional `read.title` (the artifact's stored title) to the Artifact tool's structured read result
-
Lightricks/LTX-2 v1.4.1 Patch
Official Python inference and LoRA trainer package for the LTX-2 audio–video generative model.
-
mobile-next/mobile-mcp 1.0.5 Patch +1 more
Model Context Protocol Server for Mobile Automation and Scraping (iOS, Android, Emulators, Simulators and Real Devices)
-
awslabs/mcp 2026.09.20260930084625 Rolling
fix(aws-healthomics-mcp-server): report truthful pagination state for date-filtered ListRuns by @markjschreiber in https://github.com/awslabs/mcp/pull/4639
-
paradedb/paradedb v0.26.0-rc.2 Minorpre
chore: Generate Dockerfiles for v0.25.0 by @paradedb-github-bot[bot] in https://github.com/paradedb/paradedb/pull/5731
-
Tencent/BrowserSkill dsh-plugin-v0.3.2 Patch +2 more
Let AI agents use your real, logged-in browser without interrupting your work. CLI + extension for browser automation across any shell-capable AI agent.
-
tile-ai/tilelang v0.1.15 Patch
Native Huawei Ascend 950 support** (#3308): an end-to-end NPU backend with native code generation, automatic Cube/Vector scheduling and synchronization, and mixed SIMD/SIMT programming.
-
anthropics/claude-agent-sdk-python v0.2.162 Patch
Internal/Other Changes
-
mvschwarz/openrig v0.6.2 Patch
Published:** `@openrig/[email protected]` is available on npm as `latest`.
-
Zipstack/unstract v0.192.3 Patch +1 more
[MISC] Fix flaky integration CI: rig Postgres lock limit and unmocked pg_barrier dispatch by @johnyrahul in https://github.com/Zipstack/unstract/pull/2308
-
flashinfer-ai/flashinfer v0.7.0.post1 Patch
FlashInfer: Kernel Library for LLM Serving
-
Gaurav-Gosain/tuios v0.8.2 Patch
tuios 0.8.2 adds the scratch terminal and command keys. It also brings file editing and file search in the sidebar, hints on all panes and the fixes merged since v0.8.1.
-
callstack/agent-device v0.21.18 Patch +2 more
fix(daemon): probe a live daemon again before replacing it as unreachable by @okwasniewski in https://github.com/callstack/agent-device/pull/3050
-
-
Agenta-AI/agenta v0.121.6 Patch
[feat] Add Claude Sonnet 5.5 to the Claude, Pi, prompt, and pricing catalogs by @mmabrouk in https://github.com/Agenta-AI/agenta/pull/7231
-
microsoft/mcp Azure.Mcp.Server-3.0.0-beta.48 Patch IDE extension
_Tool Improvements and URL validation_
-
agentscope-ai/QwenPaw v2.2.2-beta.4 Patchpre
Your Personal AI Assistant; easy to install, deploy on your own machine or on the cloud; supports multiple chat apps with easily extensible capabilities.
-
huggingface/datatrove v0.10.1 Patch
Freeing data processing from scripting madness by providing a set of platform-agnostic customizable pipeline processing blocks.
-
conductor-oss/conductor v3.33.0-rc10 Patchpre +1 more
Conductor is an event driven agentic workflow engine providing durable and highly resilient execution engine for applications and AI Agents
-
blazickjp/arxiv-mcp-server v0.7.3 Patch
A local MCP server for agent literature work. Original-LaTeX section reads, BibTeX from arXiv metadata, and topic watches. Papers stay on disk. Search is optional.
-
vllm-project/vllm-ascend v0.28.0.rc1 Minorpre
Community maintained hardware plugin for vLLM on Huawei Ascend
-
gastownhall/beads v1.3.1-rc.2 Patchpre
Beads - A memory upgrade for your coding agent
-
vllm-project/production-stack vllm-stack-0.1.13 Patch
The stack deployment of vLLM
-
Dao-AILab/flash-attention fa4-v4.0.0.beta33 Patchpre
[CuTe, SM100] Sparse MLA bwd: keep P normalization and dS product in fp32 by @abcdabcd987 in https://github.com/Dao-AILab/flash-attention/pull/2893
-
ByteDance-Seed/VeOmni v0.1.13 Patch
VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo
-
a2aproject/a2a-python v1.2.1 Patch
1.2.1 (2026-09-30)
-
Comfy-Org/ComfyUI_frontend v1.54.9 Patch
[backport core/1.54] fix(agent): defer the consent offer past the first-run screen instead of dropping it by @christian-byrne in https://github.com/Comfy-Org/ComfyUI_frontend/pull/19351
-
yc-software/qm vendored-pi-coding-agent-0.82.0-security.5 Releasepre
Dependency artifact only; this is not a QM product release.
-
lancedb/lancedb v0.40.0-beta.12 Patchpre
🛠 Breaking Changes
-
gpustack/gpustack v2.3.0rc2 Patchpre
A GPU cluster manager for high-performance AI model serving (vLLM, SGLang) and on-demand SSH-accessible GPU instances.
-
apache/maka v0.2.0-dev.64.20260929 Patchpre
Developer Snapshot 0.2.0-dev.64.20260929
Nothing matches this filter.