Search

Universal long-term memory for any LLM application

Codex auth helpers for pydantic-ai OpenAI Responses models.

Profile vLLM inference under RL-style rollout workloads.

M-flow builds semantic memory layers that help LLMs reason over your data with structured knowledge graphs.

A tool routing AI package using embeddings and FAISS

A modular, production-grade Retrieval-Augmented Generation library

AI skill package for Laravel 13 PHP 8 attributes - write new code and refactor old Laravel code to attribute syntax

Abstract skill interfaces for AI — delivery, testing, architecture, security, and more. Skills define WHAT; AI decides HOW.

LLM knowledge pack for Manifesto with Codex and Claude Code integration guidance

Tiny macOS CLI for syncing one shared skills library across Claude Code, Codex, and Kilo Code.

PRAMANA/1.0 — SHREEA Protocol Layers 3+4. Three strands. Four signals. Five levels. HBT PRB + kitchen_mask + TraversalResult + DomainBranchIndex types. Binary truth at every level.

OpenCode authentication plugin for OmniRoute API with /connect command and dynamic model fetching

AI-powered codeless UI & API automation framework

Zero-dependency safety rails for LLM tool calling

A general purpose LLM tool — get-me-my-meme gives your AI the power to respond with memes

Auto-generate Claude Code documentation from your actual source code — Standards, Rules, Skills, and Guides tailored to your project

Offline-first prompt improvement CLI with optional LLM mode

Build consistent Qwen Code skills using standardized syntax and guided workflows

Graveyard Protocol CLI — close empty Solana SPL token accounts and reclaim locked SOL

Authority-aware OpenAI tool integration for Baseline.