Daily Spotlight
Curated open source finds. Hand-picked for utility and relevance.
Trending & Rising
High-velocity repos and community favorites.
Trending Now
High-velocity repositories gaining rapid community traction.
BMayhew/awesome-sites-to-test-on
A curated list of sites to practice testing on
coreos/rpm-ostree
⚛📦 Hybrid image/package system with atomic upgrades and package layering
mrwadams/stride-gpt
An AI-powered threat modeling tool that leverages OpenAI's GPT models to generate threat models for a given application based on the STRIDE methodology.
Hidden Gems
High-utility projects flying under the radar. Low stars, high potential.
New Arrivals
Recently created projects with active commits and initial traction.
ErickJ3/raptor
🦖 Raptor: Realtime Abstracted Path Tree Observer
Tarunsaisrinivas/React-Cursor
✨ A collection of modern, animated cursor components Smooth, customizable, and plug-and-play. Fun to explore. Easy to integrate. Level up your UI with delightful cursors that respond to interaction. Built with modern web technologies—crafted for developers who care about the details.
zinja-coder/zin-mcp-client
MCP Client which serves as bridge between mcp servers and local LLMs running on Ollama, Created for MCP Servers Developed by Me, However other MCP Servers may run as well
Early Stage
Young, active projects building towards their first 100 stars.
Latest Picks
Handpicked repositories selected for utility and relevance.

Hermes Agent vs OpenClaw: Privacy-First Alternative
NousResearch's Hermes Agent takes a different path than OpenClaw: autonomous skill creation instead of human-authored tools, zero telemetry instead of data collection, and support for 200+ models instead of tight integration. Both approaches have merit—Hermes prioritizes privacy and self-improvement for developers who want control over their AI agents.
hermes-agent
NousResearch

Hello-Agents: Learn What No-Code AI Tools Actually Do
Dify and Coze make building AI agents accessible through drag-and-drop interfaces, but they abstract away the underlying principles. Hello-Agents fills the gap with a systematic 16-chapter tutorial that teaches agent architecture from scratch, backed by a 900-person learning community despite some documentation rough edges.
hello-agents
datawhalechina

Why gstack Hit 60K Stars in Two Weeks
gstack went from zero to 60,000 GitHub stars in two weeks, becoming a lightning rod for questions about AI development tooling. Role-based workflow structure resonated with developers tired of generic chatbots, but critics questioned whether the innovation was real or amplified by Y Combinator's CEO sharing his personal setup.
gstack
garrytan

NanoClaw: 500 Lines That Isolate AI Agents in Containers
OpenClaw runs agents in a single Node process with shared memory. NanoClaw's creator saw a security problem and built 500 lines of TypeScript that puts each agent in OS-level containers. The Docker partnership validates the isolation strategy, though running inside containers brings its own challenges.
nanoclaw
qwibitai

CLI-Anything: Making Desktop Software Agent-Native
GUI automation is fragile, APIs don't cover professional features, and naively editing project files fails when effects render at runtime. CLI-Anything solves the 'rendering gap' by generating CLIs that call real software backends—Blender for rendering, LibreOffice for exports—instead of reimplementing features or scraping pixels.
CLI-Anything
HKUDS

Beads: AI Coding Agents That Remember Past 1 Hour
AI coding agents like Claude and Cursor hit a wall after about an hour—they forget context, lose track of dependencies, and waste tokens re-analyzing problems. Beads gives them a memory system using task graphs and topological sort, extending sessions to 12+ hours according to DoltHub's testing. We cover the technical approach, real-world results, and honest trade-offs including merge conflict issues.
beads
steveyegge
