GitHub 2026-03-14 Monthly Trending Projects

RuView: edge WiFi CSI pose/vital sensing on ESP32; fits rescue and research use, but requires CSI hardware and has unclear license/community.

AIRI is a hybrid Web/local self-hosted virtual character platform focused on real-time voice, game interaction, and multi-platform deployment—suited for developers and creators seeking controllable, customizable virtual streamers and digital companions.

This repository compiles practical guidelines and orchestration patterns for Claude Code, detailing Commands, Sub-Agents, Skills, Hooks, MCP servers and workflow recommendations to help teams design agent-based workflows. It is a useful reference for architects and engineers exploring Claude Code practices. However, the lack of a declared license, unclear language stack, and zero visible contributors/releases mean production adoption requires careful legal and maintenance risk assessment.

Extensible super-agent runtime with skills, sandbox execution and memory.

OpenClaw is a local-first personal AI assistant that runs on your devices, offering multi-channel access, voice wake and a live Canvas — suited for privacy-focused users and developers seeking deep integrations.

Hugging Face Skills delivers reusable skill bundles standardized via SKILL.md, compatible with multiple coding agents and shipped with manifests and examples—useful for delegating Hub operations and training/evaluation workflows to agents—while license clarity and community activity require verification.

MiroFish is a multi-agent, long-memory simulation prediction platform that maps real-world seed data into high-fidelity digital worlds, enabling natural-language-driven scenario simulations and automated report generation for public opinion, policy and creative sandboxing, while requiring attention to licensing and operational risks.

Heretic pairs directional ablation with TPE-based automatic tuning to offer a one‑click, low‑damage decensorship tool for GPU-equipped researchers and engineers—useful for probing model semantics and generative boundaries, but requiring careful legal and misuse risk assessment.

OpenViking abstracts agent context as a manageable 'filesystem', offering tiered loading, directory-recursive retrieval and visualization to reduce retrieval cost and improve observability—suitable for AI agent systems requiring long-term memory and complex context management.

A general-purpose sandbox platform for AI applications offering multi-language SDKs, unified APIs and Docker/Kubernetes runtimes to run agents, code execution and training workloads in isolated environments for debugging and evaluation.

Claude‑skills supplies 66 specialized skills and 9 workflows for Claude‑based conversational programming, emphasizing context engineering and progressive disclosure—ideal for full‑stack teams using Claude as an expert pair programmer; however, there are no formal releases and contributor activity is low, so evaluate maintenance and compatibility risks before adoption.

Pi is a developer-focused monorepo toolkit offering a unified multi-provider LLM API, agent runtime, coding agent CLI, terminal and web UI components, and tools to manage vLLM GPU pods—suited for teams with TypeScript/Node skills building and deploying custom AI agents.

A progressive, hands-on tutorial repository for developers and researchers that demonstrates how to build Claude-like AI coding agents across five increasing-complexity examples, emphasizing clear concepts and runnable samples.

An open collection of system prompts and models for prompt engineering and AI integration, covering many agents and developer tools—useful for experimentation and reuse, but requires review of GPL licensing and compliance risks before commercial adoption.

Superpowers provides a composable skills library and TDD-enforced agent workflow for Claude Code, enabling end-to-end automation from design and planning to subagent execution, with plugin integration and extensibility.

Superset is a macOS terminal workspace designed to run arbitrary CLI coding agents in parallel with worktree isolation, agent monitoring, and built-in diff/edit capabilities to reduce context switching and speed local review workflows; however, platform/runtime dependencies and the repository's low apparent community activity warrant compatibility and maintenance risk assessment before production adoption.

Rowboat is a local-first AI coworker that accumulates long-lived context in an editable Markdown knowledge graph, supports pluggable models and external tool integrations, and targets privacy-conscious professionals focused on sustained knowledge management.