Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Create and revise irudd-plan canvases or implement a selected work item through its required authenticated MCP integration. Use when a task names an irudd-plan human link or MCP reference, or requests collaborative canvas planning. Keep implementation retrieval focused on the selected packet.
Use Jev as System One to save the user reasoning-model tokens, time, and cost on bulk reading or repeated semantic decisions. Trigger when sifting large documents or tool outputs, selecting evidence or actions from many candidates, or checking many items against criteria, even if the user never mentions Jev. Also use for "Jevify this", Jev question design, and product work where frequent, inexpensive judgments could improve the experience.
(NS) Interview to map the five-block reference architecture, decompose a workflow into subtasks, decide agent vs rule vs approval gate per subtask, assign trade-off budget per component, lock an architecture change signal, then recommend LangGraph vs CrewAI, topology, personas, tools, and models — after conceptual adaptation design from `ns-agent-engineering` when agent vs RAG vs fine-tune is not yet locked. Use when building multi-agent systems, choosing LangGraph/CrewAI, designing crews/graphs, HITL workflows, "agents working together", producing official architecture decision records, or reverse-documenting an agent that already exists. Do NOT use for conceptual Prompt/RAG/Agent/Fine-Tune ladder (`ns-agent-engineering`), general app requirements (`/ns-spec-driven` Clarify), or coding without architecture intent.
Quick-start guide and API overview for the OpenServ Ideaboard - a platform where AI agents can submit ideas, pick up work, collaborate with multiple agents, and deliver x402 payable services. Use when interacting with the Ideaboard or building agents that find and ship ideas. Read reference.md for the full API reference. Read openserv-agent-sdk and openserv-client for building and running agents.
Pi coding agent customization — Themes, Keybindings, Prompt Templates, System Prompt overrides, and Settings. Use when configuring or modifying the visual appearance, UI colors, keyboard shortcuts, prompt templates, or the default system prompt behavior of the pi terminal agent. Use for "custom theme pi", "change keybindings", "modify system prompt", "SYSTEM.md", "APPEND_SYSTEM.md", "settings.json", "keybindings.json", "prompt template", even if pi-cli-workspace is already active.
Expert in designing, optimizing, and evaluating prompts for Large Language Models. Specializes in Chain-of-Thought, ReAct, few-shot learning, and production prompt management. Use when crafting prompts, optimizing LLM outputs, or building prompt systems. Triggers include "prompt engineering", "prompt optimization", "chain of thought", "few-shot", "prompt template", "LLM prompting".
Universal prompt engineering techniques for any LLM. Use when crafting, optimizing, or reviewing prompts for AI models. Triggers on requests like “improve this prompt”, “write a system prompt”, “optimize my instructions”, “help me prompt engineer”, “audit this prompt”, “review my prompt”, or when building agentic systems that need structured prompts.
Use at a checkpoint after real friction — a task where something was repeatedly confusing, a skill misfired or failed to fire, or a workaround was invented that should not have been needed. Concerns turning friction into a change to the skills themselves, and knowing when the right change is to cut rather than add. Trigger at session end after substantial work, or when the same problem has now happened more than once. Do NOT use reflexively after routine tasks, or mid-task.
Quality-control workflow for complex, multi-step tasks that need research, clarification, a verifiable plan, independent review, and evidence-backed delivery. Use for substantial implementations, decision studies, multi-file deliverables, and exhaustive document or visual checks, or when the user explicitly asks to use SureForge. Scale down for small tasks; do not activate automatically for casual questions, simple lookups, or trivial edits.
Strip AI-slop patterns from reasoning traces (chain-of-thought, extended thinking, agent decomposition) — not final prose. Reasoning text has its own slop catalog that regular unslop doesn't target: over-explaining the question, over-hedging, over-decomposing trivial problems into 6-bullet substeps, infinite-loop rationalization. Trigger: /unslop-reasoning, "clean up my reasoning", "fix this chain of thought", "this CoT sounds robotic". Applies to reasoning output; does not override regular /unslop mode.
Reviews AI agent architectures for security risks including permission model design, least-privilege enforcement, human-in-the-loop gate placement, blast radius containment, audit trail completeness, rollback capability, and multi-agent trust boundaries. Auto-invoked when reviewing agentic AI systems where LLMs invoke tools, take autonomous actions, or operate in multi-agent configurations. Produces a structured architecture security assessment mapped to OWASP Agentic AI threats and NIST AI RMF 1.0.
Run a strict multi-iteration Reflex Loop with phases (PLAN, PRODUCE||PREPARE, EVALUATE, CRITIQUE, REFINE) to improve an artifact until quality gates pass or iteration limits are reached. Use when user asks for iterative refinement, quality-gated generation, or "generate -> critique -> refine" loops.
A workflow system for writing novels collaboratively with AI. 9 agents work together to complete the full writing flow from setting to archiving. Entry check → initialization/migration → hand off to novel-agent for scheduling. Use cases: writing a new novel from scratch, importing an existing novel.
Use when working with AWS Strands Agents SDK or Amazon Bedrock AgentCore platform for building AI agents. Provides architecture guidance, implementation patterns, deployment strategies, observability, quality evaluations, multi-agent orchestration, and MCP server integration.
Initialize a BMAD planning workspace: pick a scale-adaptive TRACK (Quick Flow / BMad Method / Enterprise) interactively, then scaffold the output folder, a config file, an empty decision-log.md, and a project-context.md "constitution". Use when the user says "initialize BMAD", "set up BMAD", "start a new BMAD project", "scaffold the planning workspace", "bmad init", "create the bmad config", or is beginning planning and has no bmad-output/ folder yet. Also use when the user asks "which track should I use?" or "what scale is my project?". This is the FIRST skill to run before any other planning workflow (brief, PRD, architecture, stories).
Delegate browser/UI verification and computer-use tasks to Codex — launching the app, clicking through flows, verifying UI/UX changes visually. Use when a change needs to be seen working in a real browser and screenshots would burn main-loop tokens.
Internal downstream skill for ctf-sandbox-orchestrator. CTF-sandbox workflow for AI-agent, prompt-injection, MCP or toolchain, cloud, container, CI/CD, and supply-chain challenges. Use when the user asks to analyze prompt-to-tool flows, retrieval poisoning, mounted secrets, deployment drift, runtime-vs-manifest mismatches, registry provenance, or CI-produced artifacts under sandbox assumptions. Use only after `$ctf-sandbox-orchestrator` has already established sandbox assumptions and routed here.
Internal downstream skill for ctf-sandbox-orchestrator. CTF-sandbox workflow for prompt-injection, retrieval poisoning, memory contamination, planner drift, MCP or tool-boundary abuse, and agent exfiltration challenges. Use when the user asks to analyze prompt injection, retrieval poisoning, memory contamination, planner drift, tool-argument corruption, or secret exposure caused by an agent chain. Use only after `$ctf-sandbox-orchestrator` has already established sandbox assumptions and routed here.
Downloads, installs, and relocates new agent skills to the global Antigravity shared directory (`~/.gemini/config/skills/`) to make them available across all tools. Use when the user requests to download, install, import, setup, or update new agent skills, or asks to relocate installed skills.
Blocks dangerous commands before execution via a Rust-based Claude Code hook. Use when configuring agent safety guards, setting up destructive command blocking, or auditing CLI protection rules. Use for git reset protection, rm -rf interception, force-push blocking, pack-based command filtering, and PreToolUse hook safety.
Connect a Feishu (Lark) bot to Clawdbot via WebSocket long-connection. No public server, domain, or ngrok required. Use when setting up Feishu/Lark as a messaging channel, troubleshooting the Feishu bridge, or managing the bridge service (start/stop/logs). Covers bot creation on Feishu Open Platform, credential setup, bridge startup, macOS launchd auto-restart, and group chat behavior tuning.
Graph of Thoughts (GoT) Controller - manages research graph state, executes graph operations (Generate, Aggregate, Refine, Score) and optimises the quality of research paths. Use this skill when the research topic is complex or multi-faceted, requires strategic exploration (depth vs breadth) or demands high-quality research.
Use when creating or developing Laravel features, before writing code or implementation plans - refines rough ideas into fully-formed Laravel designs through collaborative questioning, alternative exploration, and incremental validation.
Universal meta-skill that bootstraps ANY software project regardless of programming language. Auto-generates 15-30+ hyper-detailed, project-specific coding skills BEFORE writing code. Language-agnostic: works with TypeScript, Python, Go, Rust, Java, C#, Swift, Kotlin, PHP, Ruby, and polyglot projects. AGGRESSIVE version enforcement: EVERY technology version MUST be verified via real-time lookup — never use memorized versions. Generates skills for architecture, security, performance, privacy, testing, error handling, accessibility, observability, data modeling, API design, DevOps. Each skill enforces zero-bug standards with concrete code examples, anti-patterns, and measurable budgets. Triggers on: "bootstrap", "new project", "start project", "create project", "set up", "generate skills" — use BEFORE any code exists.
Use when you have already determined the scope of Discover (P0/P1/P2) and now need to quickly establish the Level-0 north star (memory) and the Level-1 map-layer index skeleton (components/products) of `.aisdlc/project/`, so that evidence can later be filled in module by module without double-writing or drift.
Keep a reviewable decision trail for long-running or unattended work: a TSV log with one row per decision (what, why, evidence, result). Local by default; commit it when a reviewer needs the trail to trust the result. Use for /show-me-your-work, autonomous or multi-phase runs, or work a human reviews after stepping away.
Create new skills (SKILL.md files), modify and improve existing skills, and design skill descriptions for accurate triggering. Use when the user wants to create a new skill from scratch, edit an existing skill, optimize a skill's description, or convert a workflow they just demonstrated into a reusable skill.
Use when a skill or slash command is missing, duplicated, not loading, or inconsistent between `.agents/skills` and `.claude/skills`; includes “skill 没生效”“检查 skill 链接” and broken skill links. Audits project and global skill layouts without changing them.
Quebra um plano, spec ou a conversa atual em um conjunto de tickets tracer-bullet, cada um declarando suas blocking edges, publicados no tracker configurado — edges como texto em um arquivo por ticket localmente, ou links de blocking nativos num tracker de verdade.
A structured root-cause investigation protocol for complex, ambiguous, or multi-layer technical problems. Activate this skill whenever: a problem has resisted two or more fix attempts; the root cause is unknown or assumed; you are tempted to try a variation of something that already failed; a system has multiple interacting layers (hardware, OS, runtime, middleware, config, network); the user says "ultrathink", "think deeper", "figure out why", "stop guessing", "find the root cause", or "it's still broken after your fix". Also activate proactively when you catch yourself about to write a fix before you have verified the cause — that instinct is the signal the protocol is needed. The protocol enforces three disciplines that distinguish root-cause investigation from trial-and-error: (1) explicit THOUGHT/ACTION/OBSERVATION cycles, (2) a hard gate that blocks implementation until the cause is verified by direct evidence, and (3) structured escalation when in-process diagnostic tools are exhausted.
Use when coordinating subagents for coding, research, repair, or review work where delegated workers may drift, stall, crash, duplicate, or edit a shared workspace and the coordinator must preserve scope, verification, and user-consent boundaries.
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.