Agent Skills Hub

cmux

Control macOS cmux terminal topology, workspaces, and pane layouts via CLI. Ideal for AI coding agents requiring deterministic multi-pane navigation, surface routing, and attention cues.

evaluation

Build systematic evaluation frameworks for AI agents using multi-dimensional rubrics, LLM-as-a-judge, and regression testing to measure performance, quality, and context engineering effectiveness.

prompt-writing

Create, refine, and optimize high-quality YAML prompts for AI assistants using structure guidelines, template patterns, and quality standards.

solo-content-gen

Generate a promotion content pack from PRDs or READMEs, including LinkedIn posts, Reddit drafts, and Twitter threads.

zotero-mcp-code

Efficiently search your Zotero library using Python code execution. Enables comprehensive multi-strategy queries, automated deduplication, and relevance ranking without context overflow or system crashes.

skill-generator

An intelligent generator for Claude Code Skills that automates the creation of structured prompts, YAML frontmatter, and supporting file architectures.

clawhub

ClawHub is the official registry and CLI tool for managing OpenClaw AI agent skills. Search, install, version-control, and publish custom skills to your local OpenClaw workspace.

zed-editor

Provides comprehensive knowledge on Zed Editor and the Agent Client Protocol (ACP), including AI agent integration, performance tuning, and configuration for professional development workflows.

summarize-interview

Synthesize interview transcripts into a structured template including Jobs to Be Done (JTBD), satisfaction signals, and actionable items.

freeride

Manages free AI models from OpenRouter for OpenClaw. Ranks models by quality, configures fallbacks for rate-limit handling, and updates openclaw.json automatically.

mcporter

A command-line tool to list, configure, authenticate, call, and inspect Model Context Protocol (MCP) servers via HTTP or stdio.

debug-hooks

Systematic debugging workflow for Claude Code hooks. Use to troubleshoot non-firing hooks, output errors, or unexpected behavioral issues.