Minimal educational LLM agent harness: single-agent tool-use loop with streaming tool execution and context summarization. ~450 lines of readable Python for beginners.
npx add-skill yo666666yo/minimal-agent-harnessMinimal educational LLM agent harness: single-agent tool-use loop with streaming tool execution and context summarization. ~450 lines of readable Python for beginners.
Source: https://github.com/yo666666yo/minimal-agent-harness
Discovered from GitHub repositories pushed in the last 24 hours for agent skills, Claude/Codex/Gemini workflows, MCP tooling, and adjacent AI-agent automation.
Codex skills for building and maintaining an LLM wiki knowledge base inspired by Andrej Karpathy's LLM Wiki pattern.
A set of manually-invoked Claude Code / Codex skills for taking an engineering feature from "rough idea" to "merged code, one reviewable commit per step."
MCP server that lets AI agents discover, pay for, and execute SKILL.md endpoints via the x402 protocol
The official TypeScript SDK for Model Context Protocol servers and clients
Agent skill or AI-agent tooling repository surfaced by cross-model daily skill hunt.
Multi-CLI (Claude Code + Codex + Gemini) project bootstrap with shared AGENTS.md, /wrap session skill, and CLI-first tool preference.