Today's AI Frameworks & SDKs: Fastest-Growing Projects — April 17, 2026
Today's AI Frameworks & SDKs space saw a surge in activity around tools that enable developers to build and manage complex AI systems, with a focus on decentralized computing, formal verification, and efficient infrastructure management. Several repositories gained significant traction, with growth scores exceeding 20% and star counts increasing by hundreds. Notably, many of these projects leverage cutting-edge technologies like peer-to-peer networks, Kubernetes, and tree-sitter indexing to streamline AI development.
jxnxts/mcp-brasil, with a growth score of 57.30 and 1,381 stars, is a prominent example of this trend. This repository provides an MCP server for 41 public Brazilian APIs, allowing developers to easily integrate these services into their applications. Its rapid growth can be attributed to the increasing demand for standardized API interfaces in Brazil.
Agent-FM/agentfm-core, boasting a growth score of 33.17 and 32 stars, is another noteworthy project. This peer-to-peer network enables users to run massive AI workloads across a global mesh of idle CPUs and GPUs, effectively creating a decentralized AI supercomputer. Its growing popularity stems from the need for more efficient and cost-effective ways to train large AI models.
yogthos/chiasmus, with a growth score of 30.90 and 60 stars, offers an MCP server that grants language models access to formal verification capabilities. This repository's growth can be attributed to the increasing importance of formal verification in ensuring the reliability and security of AI systems.
Tencent-Hunyuan/HY-SOAR, despite having a relatively low commit count, boasts a growth score of 29.50 and 47 stars. This project focuses on self-correction for optimal alignment and refinement in diffusion models, addressing a crucial challenge in AI research. Its growth is likely driven by the growing interest in improving the accuracy and efficiency of diffusion models.
416rehman/DeepZero, featuring a growth score of 27.25 and 89 stars, provides an automated vulnerability research framework that utilizes AI agents to analyze Windows kernel drivers for exploitable IOCTLs. This repository's popularity can be attributed to the increasing need for proactive security measures in the face of rising cyber threats.
Yuan-lab-LLM/ClawManager, with a growth score of 25.38 and 575 stars, offers a Kubernetes-native control plane for AI agent instance management, enabling governed AI access and runtime orchestration. Its growth is likely driven by the increasing adoption of Kubernetes in enterprise environments.
codefromkarl/ContextAtlas, boasting a growth score of 22.43 and 25 stars, provides context infrastructure for AI coding agents, featuring hybrid retrieval, project memory, and retrieval observability via CLI or MCP server. This repository's growth can be attributed to the growing importance of efficient context management in AI development.
OpenEnvision/Awesome-Multimodal-Modeling, with a growth score of 21.31 and 247 stars, is a curated list of resources for multimodal modeling, covering MLLM, UMM, and NMM. Its popularity stems from the increasing interest in multimodal learning and its applications.
jnMetaCode/agency-orchestrator, featuring a growth score of 19.57 and 313 stars, provides a multi-agent framework that works with existing AI subscriptions, eliminating the need for API keys. This repository's growth can be attributed to the growing demand for more flexible and accessible AI solutions.
rhino-acoustic/NeuronFS, boasting a growth score of 18.36 and 136 stars, offers an innovative file system that leverages B-tree indexing to govern AI agents efficiently. Its growth is likely driven by the need for more efficient infrastructure management in AI development environments.
jxnxts/mcp-brasil, with a growth score of 57.30 and 1,381 stars, is a prominent example of this trend. This repository provides an MCP server for 41 public Brazilian APIs, allowing developers to easily integrate these services into their applications. Its rapid growth can be attributed to the increasing demand for standardized API interfaces in Brazil.
Agent-FM/agentfm-core, boasting a growth score of 33.17 and 32 stars, is another noteworthy project. This peer-to-peer network enables users to run massive AI workloads across a global mesh of idle CPUs and GPUs, effectively creating a decentralized AI supercomputer. Its growing popularity stems from the need for more efficient and cost-effective ways to train large AI models.
yogthos/chiasmus, with a growth score of 30.90 and 60 stars, offers an MCP server that grants language models access to formal verification capabilities. This repository's growth can be attributed to the increasing importance of formal verification in ensuring the reliability and security of AI systems.
Tencent-Hunyuan/HY-SOAR, despite having a relatively low commit count, boasts a growth score of 29.50 and 47 stars. This project focuses on self-correction for optimal alignment and refinement in diffusion models, addressing a crucial challenge in AI research. Its growth is likely driven by the growing interest in improving the accuracy and efficiency of diffusion models.
416rehman/DeepZero, featuring a growth score of 27.25 and 89 stars, provides an automated vulnerability research framework that utilizes AI agents to analyze Windows kernel drivers for exploitable IOCTLs. This repository's popularity can be attributed to the increasing need for proactive security measures in the face of rising cyber threats.
Yuan-lab-LLM/ClawManager, with a growth score of 25.38 and 575 stars, offers a Kubernetes-native control plane for AI agent instance management, enabling governed AI access and runtime orchestration. Its growth is likely driven by the increasing adoption of Kubernetes in enterprise environments.
codefromkarl/ContextAtlas, boasting a growth score of 22.43 and 25 stars, provides context infrastructure for AI coding agents, featuring hybrid retrieval, project memory, and retrieval observability via CLI or MCP server. This repository's growth can be attributed to the growing importance of efficient context management in AI development.
OpenEnvision/Awesome-Multimodal-Modeling, with a growth score of 21.31 and 247 stars, is a curated list of resources for multimodal modeling, covering MLLM, UMM, and NMM. Its popularity stems from the increasing interest in multimodal learning and its applications.
jnMetaCode/agency-orchestrator, featuring a growth score of 19.57 and 313 stars, provides a multi-agent framework that works with existing AI subscriptions, eliminating the need for API keys. This repository's growth can be attributed to the growing demand for more flexible and accessible AI solutions.
rhino-acoustic/NeuronFS, boasting a growth score of 18.36 and 136 stars, offers an innovative file system that leverages B-tree indexing to govern AI agents efficiently. Its growth is likely driven by the need for more efficient infrastructure management in AI development environments.