AI solutions radar, 20 Nov 2026: dsh-Go leads 8 new projects
The most-starred open-source AI project created on GitHub in the last 30 days is ccch1mneyyy/dsh-Go (28,176 stars), which describes itself as “dsh-go — DeepSeek Harness 装进口袋:手机本机运行,也能遥控你的电脑。” Below: 7 more new projects, 2 established projects that shipped releases this week and 5 new arXiv papers.
Key takeaways
- The 8 new projects below have gathered 50,377 stars between them, all within 30 days of creation.
- The most common tag among them is “claude-code” (3 projects).
- Newest entrant: sybil-solutions/inference-optimisation-skill, created on 16 Nov 2026.
- 2 established projects shipped releases in the past week, including NousResearch/hermes-agent v0.23.2.
- Newest paper on applied AI in real organisations: “Mapping the AI Supply Chain: Dependencies, Vulnerabilities, and Market Concentration in the AI Ecosystem” (18 Nov 2026).
Rising this month
The most-starred open-source AI projects created in the last 30 days.
ccch1mneyyy/dsh-Go
★ 28,176 · Dart · AGPL-3.0 · created 25 Oct 2026
dsh-go — DeepSeek Harness 装进口袋:手机本机运行,也能遥控你的电脑。
wangzan101/Mika
★ 9,178 · Kotlin · Apache-2.0 · created 29 Oct 2026
An on-device AI agent for Android that operates apps and the web to get things done for you.
mdiri-ai/SoduWork
★ 4,191 · JavaScript · MIT · created 29 Oct 2026
The AI-native engine for teams and their agents. Multi-harness, multi-agent, self-hosted.
N1nEmAn/sapper-v2
★ 2,491 · Python · MIT · created 30 Oct 2026
The first fruit of the Nüwa tree — an autonomous AI pentest legion (634-team champion, 8h unattended, 1B tokens beat 4.5B). Not just a tool: evolvable by design. Grow your own fruit on the Nüwa Foundation…
Sandflow-ai/sand3d
★ 2,173 · HTML · Apache-2.0 · created 3 Nov 2026
3D modeling in Python, built for LLMs.
JosssphZhou/poki
★ 1,653 · Swift · AGPL-3.0 · created 5 Nov 2026
Poki uses your iPhone like you do. An open-source agent app that runs on your iPhone, reads the screen, and taps, types and swipes in real apps with your own model account.
sybil-solutions/inference-optimisation-skill
★ 1,259 · Python · MIT · created 16 Nov 2026
Agent skill for hill-climbing local LLM serving speed (decode, speculative decoding, prefill, KV, startup, TTFT) on any hardware and quantization
wici-ai/inferssd
★ 1,256 · C++ · Apache-2.0 · created 22 Oct 2026
Run models larger than your RAM or VRAM by streaming MoE experts from SSD. Built on llama.cpp.
Shipped this week
Established projects that published a release in the past seven days.
NousResearch/hermes-agent
★ 264,760 · Python · MIT
The agent that grows with you
Release v0.23.2 · 16 Nov 2026huggingface/transformers
★ 169,286 · Python · Apache-2.0
Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
Release v5.23.0 · 13 Nov 2026
New research
The newest arXiv papers on applied AI in real organisations.
Mapping the AI Supply Chain: Dependencies, Vulnerabilities, and Market Concentration in the AI Ecosystem
Sarah H. Cen, Lindsey A. Gailmard, Joohan Kim et al. · 18 Nov 2026 · arXiv 2611.24927 · cs.CY
AI systems are built across many organizations that contribute models, data, compute, capital, talent, and other assets.
Grading Executable Legal Rules Against Independent Records: Transitional Provisions and Korean SME Status
Jinhyung Bae · 18 Nov 2026 · arXiv 2611.24055 · cs.CL
Laws are increasingly turned into programs that decide cases, such as whether a firm was a small or medium enterprise (SME) on a given date, and large language models (LLMs) now write them.
A Capability Framework for Evaluating Agentic AI Platforms in Regulated Enterprises
Abbas Raza Ali, Muhammad Ajmal Siddiqui, Moona Zahid · 18 Nov 2026 · arXiv 2611.24039 · cs.AI
Agentic AI platforms-systems in which LLM agents plan, invoke tools, retain memory and coordinate to change the state of enterprise systems-are being procured faster than they can be measured.
MLE Agents: Autonomous ML Engineering in Large-scale Recommendation Systems
Matthew Li, Albert Liang, Dilip Raj · 18 Nov 2026 · arXiv 2611.23910 · cs.AI
The typical workflow for improving large-scale ranking and recommendation models is slow and tedious: an engineer forms a hypothesis, implements it as a code change, trains a model, and evaluates it both offline and online, often relying…
RePlan-VLA: Progress-Aware Replanning for Long-Horizon Vision-Language-Action Models
Apoorva Vashisth, Manav Kulshrestha, Damon Conover et al. · 18 Nov 2026 · arXiv 2611.23843 · cs.RO
Vision-Language-Action (VLA) models have emerged as a promising paradigm for general-purpose robotic manipulation, yet their performance often degrades on long-horizon tasks that require multiple interdependent manipulations.
Questions and answers
What is the most popular new open-source AI project on GitHub this week?
As of 20 November 2026, ccch1mneyyy/dsh-Go is the most-starred open-source AI project created in the last 30 days, with 28,176 stars. Its description reads: “dsh-go — DeepSeek Harness 装进口袋:手机本机运行,也能遥控你的电脑。”
Which established open-source AI projects shipped updates this week?
In the seven days to 20 November 2026: NousResearch/hermes-agent (v0.23.2, 16 Nov 2026); huggingface/transformers (v5.23.0, 13 Nov 2026).
What is the latest research on applied AI in real organisations?
The newest matching arXiv paper is “Mapping the AI Supply Chain: Dependencies, Vulnerabilities, and Market Concentration in the AI Ecosystem” by Sarah H. Cen, Lindsey A. Gailmard, Joohan Kim et al., submitted on 18 Nov 2026. AI systems are built across many organizations that contribute models, data, compute, capital, talent, and other assets.
How is the ai solutions radar compiled?
Automatically every Wednesday from the public GitHub search API and arXiv. Projects are ranked by stars, and forks, archived repositories and suspected star-farm spam are filtered out. Stars are a popularity signal, not an endorsement: review the code, licence and security posture before adopting anything.
Sources
- GitHub search: topic “llm”, created in the last 30 days
- GitHub search: topic “llm”, over 20,000 stars, active this week
- arXiv: newest papers on applied AI in real organisations
Compiled automatically for the AI Knowledge Village by Dr. Rami Shaheen. Star counts are as of 00:22 UTC on 20 November 2026 and change daily. Inclusion is not an endorsement. What is ai solutions, and why it matters.
More from the village
- Agentic GitHub
Agentic GitHub radar, 22 Nov 2026: new releases from graphify, langchain and orca
Releases this week from Graphify-Labs/graphify, langchain-ai/langchain and stablyai/orca.
- MCP servers
MCP servers radar, 21 Nov 2026: bla leads 6 new projects
qybaihe/bla (6,334 stars) tops 6 new MCP servers. Plus 3 releases and 5 new papers.
- Agent skills
Agent skills radar, 20 Nov 2026: kubai-icons leads 8 new projects
kubai087/kubai-icons (1,044 stars) tops 8 new agent skills projects. Plus 4 releases.