Reusable AI skills and capabilities for agent workflows
dusama
Smart LLM router plugin for OpenClaw — classify requests and route to the best model using your own API keys. 14-dimension scoring, <1ms classification, per-prompt/session model switching.
...morereskllm
A TypeScript library for secure LLM interactions.
GitHub Actions
Ei - Local-first AI companion with persistent personas
zbigniew-sobiecki
Squint — AST + LLM driven codebase analysis and visualization
tayers
Provides tooling for feature development of Apigee proxies using YAML, JSON and ZIP formats, and REST, MCP & CLI interfaces.
...morecrazyrabbitltc
A protocol for large scale Interplanetary Intertool Agent Context
stephschofield
Beth - A ruthless, hyper-competent AI orchestrator for GitHub Copilot multi-agent workflows
ranklabs
Next.js components for Schema.org JSON-LD (@graph) using @ranklabs/schema. Deterministic, LLM-friendly structured data.
sushanth
SochDB Node.js SDK - AI-native embedded database with concurrent multi-process support
GitHub Actions
Interactive AI coding assistant CLI powered by Nuvin core
syntheticminds
Telemetry client for sending LLM logs to Wiretap
tjtanjin
A generic LLM connector for integrating Large Language Models (LLMs) in React ChatBotify!
GitHub Actions
The session browser for coding agents — Claude Code, Codex, Kimi, Gemini CLI, OpenClaw, Cursor, Aider, and more.
acruz24100
Cost-optimized LLM execution engine for TypeScript, smart routing, prompt caching, budget guards, and pre-fetch pipelines
...moresanliy
用于接入open-llm-vtuber的插件
nan0web
LLM Protocol layer for NaN•Web — Sessions, Drivers, Models, Streaming. One Logic — Many UI.
rashed_iqbal
CLI tool and Claude Code plugin to switch between API providers (Anthropic, Kimi, Qwen, DeepSeek, MiniMax, Z.ai)
hang.yang
chaijs
Object value retrieval given a string path
GitHub Actions
The official TypeScript SDK for Memori
microsoft1es
Validate Azure REST API Specifications
adrianhdezm
This is just another SDK for the common LLM API providers.
flexatone
A lightweight wiretap for LLM SDKs: capture all requests and responses with a single line of code
marcelo_emmerich
Local test runner for multi-turn, tool-calling LLM prompts