Reusable AI skills and capabilities for agent workflows
sup3x
Eco mode for Claude Code. /eco: -31% to -73% output tokens with critical findings intact; /eco-max: up to -75% with lowered effort. Measured hardest on Claude Fable 5 (fable5), deep-studied on Sonnet 5, works on Opus 4.8 too. We publish our negative results. 82 raw benchmark runs.
...moreRoninForge
Cursor plugin for modern Android (Kotlin 2.x + Jetpack Compose + Material 3). Teaches StateFlow + collectAsStateWithLifecycle, Hilt, type-safe Navigation Compose, KSP, Version Catalogs, kotlin-compose plugin. Catches 20 legacy regressions LLMs still produce.
...moreGitHub Actions
Corti CLI — wire Corti as an OpenAI-compatible provider into terminal AI coding agents (OpenCode, Pi, Crush, ForgeCode, Codex CLI) and more.
...moreliveedevteam
Install the nextjs-cicd Claude Code skill via npx
joktec_dev
Hybrid agent skill pack for using @joktec/* libraries in consumer projects.
ivandudas
n8n community nodes for the Circus AI Workflow Orchestration Platform
ladyhunterbear
Shared TS->WASM build, sandbox, and validate pipeline for atheory-ai plugin SDKs (ce-plugin-sdk, argent-plugin-sdk).
lionhummer
Heyo sandbox provider for the Vercel AI SDK. Run agent-generated commands and code in isolated Heyo microVMs.
GitHub Actions
A Node.js vision-model toolkit: detection/tracking with zone triggers, visual question answering, captioning, and image embeddings (joint and vision-only), importable directly with no HTTP required.
...morelk-hour
AI-powered project-aware developer assistant. Huno knows your codebase.
GitHub Actions
Project-level distribution kit for Project Governance System.
shekhardhangar
A verifier-centered, prompt-driven engineering workflow for pi: set an executable target, then an unattended verify-fix loop runs against it — on a branch, fenced by a guardrail hook, with a deterministic verifier that can't be faked or weakened.
...moreshekhardhangar
A verifier-centered, prompt-driven engineering workflow for pi: set an executable target, then an unattended verify-fix loop runs against it — on a branch, fenced by a guardrail hook, with a deterministic verifier that can't be faked or weakened.
...morephdthu26
Interactive REPL shell for exploring LanceDB databases
royvillasana
Spec-Driven Development for Design Engineers — installs the Claude Code skill toolkit into any front-end project
jacques_gordon
Expo module for Mapbox Navigation SDK with 16KB page size support, NDK27, and Mapbox Maps v11.11.0+
GitHub Actions
Orionfold Relay — a local-first, multi-agent orchestration runtime and builder scaffold for AI-native work.
hemacharanpyla
∞ Infinity CLI — terminal AI chat with real-time response timing
tejaswi.s
Govern AI coding agents (Claude Code, Cursor, Codex, GitHub Copilot CLI) from a hook. It blocks dangerous tool calls locally and offline (destructive commands, secret exfiltration, prompt-injected tool results), then optionally reports the session to Axio
...moredankat
Cross-session memory layer for the Vercel eve agent framework
GitHub Actions
A Claude Code statusLine wrapper that bridges context % and rate-limit telemetry to your hooks and plugins, while keeping your existing bar.
...morehdariodev
Comprehension-Driven Development — make sure you understand the code the AI wrote before it merges. A comprehension gate for AI-assisted coding: per-commit micro-checks, COMPREHENSION_DEBT tracking, and a CI audit. Zero dependencies.
...morewanxunyang
终端编码 agent:LLM + tool-call 循环 + 流式输出(含思考)+ 16 个工具,接任意 OpenAI 兼容后端。
deftio
A standalone, cross-platform desktop markdown editor with rich rendering and LLM chat integration