Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Unity of knowledge and action: proactively identify real development work that is starting, in progress, or just finished; run an evidence-based pre-check before the first write to code, Git, or an external system; cite only confirmed constraints during execution; and after completion, once Git/test/MR evidence is obtained, automatically run a retrospective and safe evolution. Development intents in natural language — implementation, modification, debugging, refactoring, fixing, building, development verification, committing code, and Skill feature changes — should implicitly invoke it even if $development-brain is not written; pure product analysis, formal QA, release and deployment, concept explanations, ordinary Q&A, and read-only assessments of whether to update a Skill do not trigger it.
A multi-AI cross-deliberation workflow. Performs cross-review of code changes across different AI tools (Claude, Codex, Gemini, etc.). Use when: the user invokes /discuss-multi-ai, mentions wanting another AI to review changes, or pastes a deliberation request containing a commit hash. It should also trigger if the user merely says "write me the deliberation text for another AI" or "hand this change to another AI for a look".
Used to extract and generate domain-specific terminology, industry terms, internal and product terms, repository implementation maps, technical stack, and Mermaid architecture diagrams for onboarding from a specified repository, multiple repositories, or a GitHub organization. Triggered when the user requests "create a glossary", "create a domain dictionary", "make onboarding material", "review a repo/org and compile terminology", or "create a knowledge base that the AI doesn't need to re-check".
Elite reasoning and coding discipline distilled from Claude Fable 5's problem-solving patterns. Load at the START of any non-trivial task — debugging, implementing features, refactoring, architecture decisions, code review, or multi-step investigations. Raises reasoning quality by enforcing hypothesis-driven thinking, evidence-based claims, calibrated effort, root-cause fixes, and adversarial self-verification. Triggers on - "debug", "fix", "implement", "build", "refactor", "why is this failing", "design", "investigate", or any task where being wrong is expensive.
Diagnose a codebase's frontend design maturity and route the work through the right sequence of frontend-design skills. Use when the user makes a broad or vague design request — 'make this look better', 'polish the UI', 'this feels like an MVP', 'our frontend is a mess', 'where do I start with design', 'make it feel professional', 'plan a design system' — or when a request spans multiple design concerns and it's unclear which single skill applies. Assesses maturity by inspecting the actual code (Tailwind config, CSS variables, component structure, hex-code drift), picks one of five implementation paths, and produces a design roadmap: an evidence-backed maturity diagnosis plus an ordered skill sequence with the artifact each step passes to the next. If the request clearly maps to one concern (only spacing, only color, only type), skip this and invoke that skill directly.
Use when the user invokes /gauntlet-loop or asks to turn an objective, plan, spec, PRD, or requirements document (inline text, a .md file, or an .html file) into an orchestrated execution system prompt — e.g. "gauntlet this plan", "make an orchestrator prompt for this spec", "generate the execution prompt for @plan.md".
Consult Google Gemini CLI for second opinions on architecture, debugging, and security audits. Use Gemini's 1M+ context window for comprehensive code analysis. Compare Flash (fast) vs Pro (thorough) vs 3-Pro-Preview (cutting-edge). Use when: need second opinion on architectural decisions, stuck debugging after 2+ attempts, writing security-sensitive code, planning refactors (5+ files), approaching 70%+ context capacity, unfamiliar with tech stack, need peer review, or want Flash vs Pro vs 3-Pro-Preview comparison. Keywords: gemini-cli, google gemini, gemini command line, second opinion, model comparison, gemini-2.5-flash, gemini-2.5-pro, gemini-3-pro-preview, architectural decisions, debugging assistant, code review gemini, security audit gemini, 1M context window, AI pair programming, gemini consultation, flash vs pro, AI-to-AI prompting, peer review, codebase analysis, gemini CLI tool, shell gemini, command line AI assistant, gemini architecture advice, gemini debug help, gemini security scan, gemini code compare
[Hyper] Create, enter, list, remove, clean up, or repair Git worktrees for isolated branches and parallel agent sessions, including direct `git-worktree ARGUMENT` creation without follow-up questions. Use when the user asks for git worktree setup/removal, branch-per-folder workflows, parallel Codex/Claude/Cursor workspaces, or the repository-local `.hyper/git-worktree/` convention; when creating and no argument/task is clear, ask what work will happen there in the user's language, derive the folder name, then move subsequent work into the new worktree.
A hands-on skill for operating Hermes Agent. Use it when the user mentions "operating hermes", "troubleshooting hermes issues", "gateway anomalies", "platform disconnects", "cron not executing", "messy profile configuration", "skills or tools not taking effect", or "wanting another agent to help maintain hermes". Based on the hermes-agent source code and official command entry points, it prioritizes the four steps of check, locate, fix, and verify, and does not make blind changes based on guesswork.
Create, scaffold, implement, review, document, and improve Hexabot v3 actions, including custom workflow actions, external API integrations, LLM-powered actions, memory reads/writes, settings or credential handling, action bindings, extension/plugin action packages, and input/output contract design. Use when wrapping business integration requirements as Hexabot actions. Do not use for generic NestJS, frontend-only work, generic workflow YAML unless action implementation is required, chatbot copywriting, generic API clients not wrapped as Hexabot actions, or non-Hexabot automation frameworks.
Frictionless setup. Detects missing daemon, installs it, configures local memory, and verifies the full plugin → MCP → daemon round-trip. Run after `/plugin install origin@7xuanlu`, or any time the user says "set up origin", "is origin working", "fix origin".
Transform vague ideas into implementation-ready specifications through structured interviewing. Use when user describes a new feature/product idea, has a problem to solve, or needs to document requirements. Produces intent.md (technical spec for code agents) and overview.md (human-friendly summary).
Before starting a Ralph RLM loop, interview the user to sharpen the goal, stress-test the design, and write an authored PLAN.md (goal, definition of done, milestones, open questions, invariants) that the loop can follow. Use when a user wants to plan, scope, or stress-test work before delegating it to the supervisor / start_loop, or whenever PLAN.md is still the placeholder bootstrap.
Consult other LLMs (GPT-4.1, o4-mini, Gemini 2.5 Pro, Claude Opus) for second opinions on complex bugs, hard problems, planning, and architecture decisions. Use proactively when stuck for 15+ minutes or facing complex debugging. Use when user says 'ask Gemini/GPT/Claude about X' or 'get a second opinion'.
Parallel multi-AI cross-validation research workflow (major version). Dispatch N internal sub-agents + grok + gemini in parallel, automatically cross-validate findings, tier by confidence (strong consensus / partial / conflict / insufficient), generate tiered action items with arbitration. Use when user says "multi-ai research", "cross-validate", "independent consensus", "three-brain research", "multi-ai research", "parallel research", "cross-validate", or needs deep research that benefits from internal data + external 2026 consensus. NOT for quick factual Q&A, pure code reasoning, or tasks needing deep project context.
Named Tmux Manager - Multi-agent orchestration for Claude Code, Codex, and Gemini in tiled tmux panes. Visual dashboards, command palette, context rotation, robot mode API, work assignment, safety system. Go CLI.
Executes council queries by running the query pipeline across selected AI providers (Gemini, OpenAI, Grok, Perplexity), displaying formatted responses verbatim, and generating a synthesis of consensus, divergence, and recommendations. Invoked by the ask command during standard (non-agent) council queries.
Approval-gated, repo-aware requirements/spec authoring for new recursive-mode runs. Use when the user wants help creating a plan, spec, scope, or requirements for work that does not yet have a run, especially prompts like "create a plan", "help me plan", "create a spec", "write requirements", or similar derivatives.
Recommend AI short drama, micro drama and manju (漫剧) skills from a hand-checked list of 90+ projects, matched to the user's stage (novel adaptation, episode outlines and hooks, script writing, review, storyboards, character consistency, video prompts, production), video model (Seedance/即梦, Kling/可灵, MiniMax, Wan, Veo, Sora, ComfyUI) and agent (Claude Code, Codex, OpenClaw/龙虾, Qwen, Doubao, WorkBuddy). Use when someone asks which skill or tool to use for an AI short drama or 短剧/漫剧, or how to install one. Helps you choose from a manually verified list of short-drama and manju skills: novel adaptation, episode outlines and hooks, scripts, review, storyboards, character consistency, video prompts, and video production.
Analyze skill execution traces to identify issues and automatically evolve/improve skills. Use when users provide trace files (JSON) from skill runs and want to improve skill performance based on real execution data. Triggers on requests like "analyze traces", "evolve skill based on traces", "improve skill from execution history", "find issues in skill traces", or when working with skill trace/log files.
Orchestrate subagent workflows for complex tasks that benefit from decomposition, role-based delegation, and parallel execution. Use when Codex should assemble a temporary team of subagents, choose roles from a reusable role library, create a controlled fallback role when no preset role fits, coordinate read-heavy work in parallel, or handle write-heavy work with ownership boundaries, staged execution, and an integrator-led merge path.
Refresh and display the project roadmap and status dashboard. Regenerates ROADMAP.md from the artifact graph (quadrant chart, Eisenhower tables, Gantt timeline, dependency graph), creates it if missing, and opens it for review. Also serves as the project status dashboard — shows active epics, progress, actionable next steps, blocked items, and recommendations. Use when the user says 'roadmap', 'show roadmap', 'refresh roadmap', 'status', 'dashboard', 'what's next', 'overview', 'where are we', 'what should I work on', 'show me priorities', 'priority matrix', or 'what does the roadmap look like'. Also use when any skill needs to ensure ROADMAP.md is fresh before consuming it.
Configure Tavus CVI personas with custom LLMs, TTS engines, perception, and turn-taking. Use when customizing AI behavior, bringing your own LLM, configuring voice/TTS, enabling vision with Raven, or tuning conversation flow with Sparrow.
This skill should be used when the user asks to “팀 구성해줘”, “team assemble”, “전문가 팀으로 해줘”, “팀으로 해줘”, “swarm”, “병렬로 전문가 팀”, or wants to decompose a complex task into specialist roles executed via TeamCreate. Also triggers when the user describes a task clearly benefiting from parallel expert execution.
Exhaustive planning interrogation for turning vague or complex requests into precise, implementation-ready plans. Use when the user invokes $ultraplan, says "ultraplan", "ultra plan", "deep plan", "plan mode", "ask me everything", "interrogate this", or asks the agent to resolve all requirements, use cases, edge cases, tradeoffs, and open questions before implementation.
Load or manage Product Design's saved user context. Use when the user asks to set up Product Design, get started, onboard, save product or design sources, see what Product Design remembers, update saved context, or remember Product Design preferences. Examples include product URLs, Figma files, screenshots, reference images, codebase paths, Storybook, tokens, design systems, brand assets, and general product/design notes.
Use the jev CLI when a script, pipeline, CI job, or agent workflow needs a fast typed semantic judgment instead of generated text: routing or classifying text to predefined labels, yes/no gating with calibrated probability, rating along ordered levels, ranking candidate texts, or checking a condition that would otherwise require a fragile prompt-and-parse step. Also use whenever the user mentions Jev, TypeSafe, System One, calibrated judgments, or semantic gates in automation. Do not use it for text generation, multi-step reasoning, or open-ended questions.
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.