Agent Skills · AI Agent
Curated skills for Claude, Cursor and other AI agents. Copy one prompt to install.
An iterative refiner for copy, code, plans, and prompts that makes one anchored edit per round, keeps or rolls back changes, and supports quick mode or workspace recovery.
Kunlun Academy is an AI Agent skill that can be installed and invoked in compatible environments via the provided prompt.
Optimizes draft prompts using four strategies—detail expansion, role-play, structured breakdown, and reverse refinement—while asking clarifying questions when key context is missing.
Lets a single AI agent be reached and reach others on Siobac, handling messaging, approvals, and conversations.
Creates and reviews OpenCode skills and agents for design quality, delegating to the skill-review subagent and enforcing red-team security checks.
Provides a security-first vetting workflow for AI agent skills, covering source checks, code red flags, permission scope, and risk tiers.
A product discovery workflow for indie developers covering opportunity scouting, competitor teardown, MVP scope, and build planning.
Turns natural-language phone tasks into TeddyMobile Vox custom outbound calls with trial/formal modes, field completion, voice selection, and safety checks.
A technical guide for engineering agent memory and retrieval, covering short/long-term, episodic, semantic, and procedural memory, chunking, vector DB selection, and common pitfalls.
Decomposes user scenarios into keywords, searches the public SkillHub API, ranks relevant results, and outputs three to five reasoned recommendations before installing the selected one.
An AI/ML design and evaluation expert for AI Agents, producing testable, production-ready specs for prompts, RAG, guardrails, evaluation, and cost.
Match prompt frameworks to task type and experience level, then generate fill-in templates and recommend models and advanced techniques.