A concise, four‑principle guideline (as CLAUDE.md or a Claude Code plugin) that teaches LLMs to: state assumptions, prefer simple solutions, make surgical edits, and use testable success criteria — reducing overcomplication and unwanted changes when an LLM edits code.
AI Coding2026
Improves Claude Code's coding behavior with a single CLAUDE.md that prescribes four practical rules—Think Before Coding, Simplicity First, Surgical Changes, and Goal-Driven Execution—to reduce LLM assumptions, overengineering, and unrelated edits.
Provides a systematic, project-driven tutorial and runnable codebase for building AI agents, RAG pipelines, and multi-agent systems—focused on Python, LangChain/LangGraph, tooling, deployment, and an interview question bank for engineers aiming to ship production agent applications.
AI Agent2026
Generates protocol-bound GEP prompts that guide iterative evolution of AI agent behavior from runtime logs, producing auditable EvolutionEvents and reusable Genes/Capsules. Node.js-based and works offline; optional EvoMap network integration enables skill sharing, worker pools and leaderboards while git provides rollback and validation.
AI Dataset2026
Cleaned reasoning dataset of problem→thinking→solution triplets derived from Opus 4.6, provided in Parquet with ~2,160 cleaned rows (original 3,305). Filters remove empty/short/refusal/non‑substantive responses; hosted on Hugging Face under Apache‑2.0.
AI Agent2026
Provides portable agent 'skills' that steer code-generating agents toward higher-quality UI: stronger layout, typography, spacing and image-reference boards. Ships adjustable dials for design variance, motion and density and image→code pipelines for agent-led frontends.
AI Video2026
Converts scene intent into production-ready Seedance 2.0 prompts, reference-role mappings, and IP-safe rewrites for multimodal (text/image/audio/video) video generation. Ships as a modular agent-skill OS with multilingual examples, troubleshooting tools, and pro filmmaker handoff artifacts.
AI Agent2026
Provides a suite of Claude Code skills that guide the full academic pipeline—research, write, review, revise, and finalize—while enforcing integrity gates (citation verification, anti-hallucination checks) and keeping a human-in-the-loop workflow.
AI Agent2026
Convenes 18 deliberately polarized AI personas to produce structured, multi-round deliberations on hard questions across multiple LLM providers. Key features: multi-provider auto-routing, enforced dissent/novelty rules, triad/panel modes and CLI integration for Claude Code/Codex. Good for high-stakes product, strategy, or safety decisions.
AI Agent2026
Automatically evolves Hermes Agent skills, prompts, tool descriptions and code using DSPy + GEPA — mutating text via API calls, evaluating trace-based failures, and selecting variants that pass tests and human PR review. No GPU training required; runs cost roughly $2–$10 per optimization.
AI Dataset2026
JSONL dataset of Claude Opus 4.6 chain-of-thought traces paired with high-difficulty math and logic problems for supervised fine-tuning and distillation; exposes step-by-step reasoning to teach process-oriented problem solving and improve math/logic accuracy in smaller LLMs.
AI Agent2026
Provides modular “skills” that help designers and engineers audit, find, and improve UI animations and interaction decisions — optimized for use by AI agents or human-in-the-loop workflows. Distills domain-expert rules into actionable SKILL.md modules (review, improve, find opportunities, vocabulary).