Trending GitHub projects in May 2026
121 projects trended · 0 new reviews · 6 hot on Hacker News
This month's GitHub is dominated by skills, specs and memory layers that make coding agents plan and review before writing code.
- Agent skills & discipline: multica-ai/andrej-karpathy-skills, mattpocock/skills and obra/superpowers hand coding agents rule sets and skill libraries for asking first, planning, TDD and review; anthropics/claude-plugins-official hosts a vetted plugin marketplace.
- Code knowledge graphs: colbymchenry/codegraph pre-indexes a code knowledge graph that stays in sync for Claude Code, Codex, Cursor and others, while Lum1104/Understand-Anything turns a codebase into an interactive graph you can search and question.
- Agent runtime & tooling: tinyhumansai/openhuman runs many cheap LLM agents on one small server; HKUDS/CLI-Anything generates command-line harnesses so agents can drive GUI-only desktop apps; github/spec-kit adds a spec-first workflow.
- Learning & applied research: rohitg00/ai-engineering-from-scratch offers a 542-lesson curriculum on model internals, retrieval and agent runtimes; K-Dense-AI/scientific-agent-skills gives agents bioinformatics and drug-discovery skills; ruvnet/RuView senses presence and vital signs from WiFi CSI.
Top 20 of the month
Ranked by stars gained on the daily board over the month.
- 1colbymchenry/codegraph+26,168
Pre-indexed code knowledge graph for Claude Code, Codex, Cursor, OpenCode, and Hermes Agent — fewer tokens, fewer tool calls, 100% local Review coming soon
- 2tinyhumansai/openhuman+26,114
A Rust agent harness aimed at running hundreds of LLM agents cheaply on one small server, usable as a library or desktop app.
- 3
A CLAUDE.md rule set for coding agents that pushes them to ask before assuming, keep diffs small, and work toward verifiable goals.
- 4Lum1104/Understand-Anything+17,518
Graphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more. Review coming soon
- 5mattpocock/skills+16,516
Composable agent skills that add structured planning, TDD, debugging and review discipline to coding agents.
- 6obra/superpowers+13,996
A skill-based methodology that pushes coding agents to spec, plan, and review before writing code.
- 7CloakHQ/CloakBrowser+12,953
Stealth Chromium that passes every bot detection test. Drop-in Playwright replacement with source-level fingerprint patches. 30/30 tests passed. Review coming soon
- 8
A 542-lesson curriculum that teaches you to implement, test and evaluate model internals, retrieval and agent runtimes yourself.
- 9rohitg00/agentmemory+11,274
#1 Persistent memory for AI coding agents based on real-world benchmarks Review coming soon
- 10
Academic Research Skills for Claude Code: research → write → review → revise → finalize Review coming soon
- 11ruvnet/RuView+8,004
WiFi CSI sensing that estimates presence, pose and vital signs from ESP32 nodes without cameras or wearables.
- 12
A curated plugin marketplace for Claude Code, giving developers a vetted place to find and install extensions instead of hunting GitHub.
- 13supertone-inc/supertonic+4,964
Lightning-Fast, On-Device, Multilingual TTS — running natively via ONNX. Review coming soon
- 14
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables. Review coming soon
- 15HKUDS/CLI-Anything+4,637
Generated command-line harnesses that let AI agents drive desktop apps that only have a GUI.
- 16
A collection of 177 installable skills that let a general coding agent perform bioinformatics, chemistry and drug-discovery tasks.
- 17affaan-m/ECC+3,964
A shared layer of skills, memory and guardrails that makes coding agents behave consistently across different CLI harnesses.
- 18github/spec-kit+3,771
Gives AI coding agents a repeatable spec-first process so features, bug fixes and ideas move through reviewable steps.
- 19
Role-specific Claude Cowork plugins that give non-engineers prebuilt skills, commands and tool connectors for their job function.
- 20
Leaders by topic
Developer Tools
A CLAUDE.md rule set for coding agents that pushes them to ask before assuming, keep diffs small, and work toward verifiable goals.
- mattpocock/skills+16,516
Composable agent skills that add structured planning, TDD, debugging and review discipline to coding agents.
- obra/superpowers+13,996
A skill-based methodology that pushes coding agents to spec, plan, and review before writing code.
Machine Learning
- ggml-org/llama.cpp+1,250
Runs LLM and VLM inference locally across CPUs, GPUs and phones without a Python stack.
AI Engineering
- tinyhumansai/openhuman+26,114
A Rust agent harness aimed at running hundreds of LLM agents cheaply on one small server, usable as a library or desktop app.
- HKUDS/CLI-Anything+4,637
Generated command-line harnesses that let AI agents drive desktop apps that only have a GUI.
A collection of 177 installable skills that let a general coding agent perform bioinformatics, chemistry and drug-discovery tasks.
DevOps & Self-hosting
A lightweight agent that collects, processes and ships metrics and logs using 300+ plugins.
- knadh/listmonk+464
Self-host newsletter and mailing-list campaigns with segmentation, templates and multi-SMTP sending from one binary.
- medusajs/medusa+411
Open-source commerce backend with customizable modules, for teams building B2B or DTC stores without writing core commerce logic.
Security & Privacy
An autonomous agent that reads your web app's source, then runs real exploits against a staging copy to prove vulnerabilities.
Productivity & Utilities
Role-specific Claude Cowork plugins that give non-engineers prebuilt skills, commands and tool connectors for their job function.
- calcom/cal.diy+858
A fully MIT-licensed, self-hosted fork of Cal.com for people who want their own booking page without enterprise code.
Self-hosted document archive that OCRs scans and makes paper records searchable with tags, correspondents and metadata.
Learning Resources
A 542-lesson curriculum that teaches you to implement, test and evaluate model internals, retrieval and agent runtimes yourself.
- datawhalechina/easy-vibe+2,262
A hands-on course that takes non-programmers from a first idea to a shipped AI-built product.
Turns a topic or uploaded material into an interactive AI-taught classroom with slides, quizzes and simulations.
Hardware & IoT
- ruvnet/RuView+8,004
WiFi CSI sensing that estimates presence, pose and vital signs from ESP32 nodes without cameras or wearables.
Finance
A native desktop finance terminal for retail and academic analysts who want Bloomberg-style research without a Bloomberg seat.
- shiyu-coder/Kronos+1,306
A pretrained foundation model that forecasts K-line price and volume sequences for quantitative researchers and traders.