Browse agent skills

Turn an agent's plan into a versioned artifact you can annotate, lock, and execute.

make beautiful 3d websites in one skill, plus 50+ website prompts for you to use.

CLI and interactive Ink TUI for documents.js: every docx/pptx/odt/odp/ods/odg/odf/pdf/odm/odb/xlsx/markdown conversion, bridge, and editor as a scriptable command or a terminal app.

Neuralgush — local AI assistant platform. Name reservation for the forthcoming project; the real package will be published here.

Dual-Model Cooperative Inference Framework — local + cloud with Gate routing and Interrupter monitoring

Full-screen live view of every Claude Code session, its git state, and your GitHub CI

Eyes for text-only LLMs — see images (Qwen3-VL-Plus, 10-dimension description) and parse documents (MinerU precision, tables/formulas preserved) from the command line. No MCP, no Claude Code dependency.

Predicts your next message after the assistant replies in OpenCode — accept it with the down arrow

Detect when a third-party webhook payload changes shape — before it silently breaks your handler.

Initialize @abdo12k/assistant-kit in a project.

Provider-neutral AI assistant runtime with an optional portfolio-matched UI and sprite-pack support.

One brain for every AI coding agent — shared rules and skills with zero file duplication

Open-source collaborative browser editors for documents, Markdown, spreadsheets, presentations, and PDFs.

Connect Claude Code and Codex to the UVA AI Gateway with your Entra identity. No API keys on your machine.

One command turns any repository into a place where AI coding agents build software the right way: doctrine, golden paths, generated agent instructions, and machine enforced gates for FastAPI, React, TypeScript, containers, AWS CDK, and GitHub Actions

Zero-dependency, identity-based dynamic matchmaking and rules evaluation engine.

opencode 插件：在 TUI 右侧栏显示当前使用的模型供应商套餐余量

Routes each prompt to the right discipline and injects only the guidance that matches. Costs nothing on conversational turns, and surfaces the skill you forgot you installed.

Hermes aiduMEM: Advanced High-Performance Memory System with Multi-Tier Recall, FTS5 & Qdrant Engine

Focused sub-agent delegation for pi: explore / worker / reviewer agents in isolated context, with proactive dispatch injection and per-agent model selection.

Browser running for Vitest using playwright (with updated dependencies)

Turn a supported ChatGPT/Codex OAuth sign-in into a private OpenAI-compatible endpoint, starting with self-hosted n8n.

A Claude Skill that stress-tests your startup idea as a skeptical VC, then rebuilds it into a business plan with a defensible moat, MVP scope, GTM and kill criteria.

Agentic QA — scenarios written as intent, replayed on the web today and on mobile next

Technical and house-style English lint engine, CLI, and host adapters

Type-safe Result pattern and domain errors

Agent-led TypeScript repository setup and curated portable skills.

npx launcher for flexx: an agentic Solidity auditor that proves every finding with an executable Foundry PoC. Wraps the Python CLI (requires Python 3.11+ and Foundry).

Reproducible AI-engineer multi-agent teams for opencode, built on oh-my-opencode-slim

Audit every AI coding tool's instruction files in one run — Claude Code skills/agents/commands, Codex prompts and AGENTS.md, Cursor rules. What they cost every session, which never fire, and whether any are doing something they shouldn't. Deterministic, o