mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when…
Skills
Find an Agent Skill for your task. Read the guide and copy the install command.
Updated 11-05 00:16
Agents & models
25,663 results
prompt-engineering-expert
Apply prompt engineering best practices to write, refine, and optimize system prompts, user prompts, and agent instructions. Use this skill whenever the user…
prompt-evaluator
Evaluate and score user-written LLM prompts on a 100-point scale across 5 axes (Clarity, Structure, Information Content, Specificity, Context), providing…
skill-creator
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an…
piracy-handler
Piracy detection and post-processing orchestrator. Use when the wechat-search-collector comprehensive page task is complete, and piracy detection needs to be performed based on capture_results, subtasks created, a webhook plan written, or webhook dispatch/reconcile executed after a subtask reaches its terminal state. Supports sqlite/supabase…
obsidian-wiki-ingest
Use this whenever the user wants to ingest, import, index, summarize, organize, or turn Obsidian notes or external folders into an Obsidian LLM Wiki. Trigger…
obsidian-wiki-maintain
Use this whenever the user wants to repair, fix, patch, update, or apply approved structural fixes to an Obsidian LLM Wiki. Route diagnosis, validation,…
obsidian-wiki-runtime
Shared deterministic runtime dependency for the Obsidian LLM Wiki workflow skills. Use only when another obsidian-wiki skill needs root resolution or Doctor…
self-improving
Turn verified task experience into durable agent improvement by creating or refining reusable skills, while avoiding noisy or speculative updates. Use after…
sprint
Work on a batch of GitHub issues in parallel using Agent Teams. Creates one worktree per issue with TDD enforcement, coordinates via a lead agent, then…
transformers-js
Run state-of-the-art machine learning models directly in JavaScript. `@huggingface/transformers` supports text, vision, audio, and multimodal tasks in browsers…
agent-sandbox
Use this skill when designing, deploying, or operating the kubernetes-sigs agent-sandbox operator, authoring Sandbox / SandboxTemplate / SandboxClaim /…
ai
Use this skill when building AI features, integrating LLMs, implementing RAG, working with embeddings, deploying ML models, or doing data science. Activates on…
swiftui-voice-first-agent-chat
Build a voice-first agent chat in SwiftUI: a dock with a hold-to-talk mic pill, a speech-reactive glass orb overlay ('Listening…' + live caption), and…
iblai-agent-chat-sidebar
Wrap the Chat surface with the SDK's AppSidebar — projects dropdown, pinned/recent messages, and host-supplied content/footer menu items
iblai-agent-voice
Add the agent Voice tab (pick the agent's voice and configure voice calls) to your Next.js app
iblai-local-llm
Use when adding on-device LLM inference (Ollama backend) to a vibe Next.js + Tauri app. Documents the contract a vibe app must implement so the SDK can route…
chatgpt-review
Use when preparing a local ChatGPT review package or, only with explicit authorization, running an external review round whose findings must be verified…
reflection
Build a candid, evidence-grounded self-portrait from the user's own corpus — Claude Code transcripts plus their Obsidian vault. Answers 22 questions about…
review-model-performance
Run task evals across multiple Claude models and compare results side-by-side. Use when you want to understand how a skill performs across different models,…
improve-codex
Audit a codebase with the improve skill, then implement the resulting plans with OpenAI Codex (`codex exec`) running as sandboxed, browser-free, CPU-capped…
bridge
Connect this engine to another iii engine over a long-lived WebSocket so functions call across the boundary. Wire stable ids with `forward:`/`expose:`;…
devin
Drive Devin over the iii bus. Run the local Devin CLI agent (SWE-1.6) with the iii runtime context so it discovers and operates your mesh, or start and steer…
worktree
Mint isolated git worktrees for parallel agent work, track ownership in a cross-agent registry, and land finished branches back onto a target with a queued…