Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Route obsidian-mind work into the right mode: install/bootstrap the vault (ShardMind vs git clone), the daily standup/dump/wrap-up session loop, capture routes for decisions, incidents, 1:1s, and wins, the performance-graph review flow (/om-review-brief, /om-self-review, /om-peer-scan), vault maintenance and /om-vault-upgrade migration, multi-agent wiring across Claude Code, Codex CLI, and Gemini CLI, and optional QMD semantic search. obsidian-mind is a specific ready-made Obsidian vault template giving coding agents persistent, session-spanning memory through five lifecycle hooks, /om-* commands, subagents, and a competency graph — not a generic vault-building or wiki-authoring workflow. Triggers on: obsidian-mind, om-standup, om-dump, om-wrap-up, om-review-brief, om-self-review, om-peer-scan, om-vault-upgrade, brag doc, North Star.md, performance graph, competency notes, session lifecycle hooks, shardmind install obsidian-mind.
Operate Orca-managed worktrees, folder contexts, terminals, repos, automations, artifacts, skill sharing, worktree comments, and Orca's embedded browser through the `orca` CLI. Use when the user says "$orca-cli", "Orca worktree", "child worktree", "spawn codex/claude in a worktree", "read/wait/send Orca terminal", "handoff" / "handover" / "give this to another agent", "Orca browser", "orca artifacts", or "share skills". Prefer it over raw git worktree, ad hoc PTYs, or Computer Use when Orca state is involved. Use Computer Use only when a visible window needs GUI control that a CLI, filesystem, or API cannot do.
Use the public `orca` CLI to operate Orca-managed worktrees, folder contexts, terminals, repos, automations, artifacts, skill sharing, worktree comments, and the browser embedded inside the Orca app. Use when the user says "$orca-cli", "use orca cli", "Orca worktree", "child worktree", "cardStatus", "spawn codex/claude in a worktree", "read/wait/send Orca terminal", "terminal send", "full handoff", "handover", "give this to another agent", "another worktree", "Orca browser", "orca artifacts", "share HTML/Markdown", "public artifact link", "share skills", or "control the browser inside Orca". Prefer this over raw `git worktree`, ad hoc PTYs, Playwright, or Computer Use when the task touches Orca-managed state. Use Computer Use for external browser windows, webviews, or desktop UI only when the task requires OS/window-level control such as focus, menus, dialogs, coordinates, or screenshots. Use `orca-cli` for Orca's embedded pages and a page-automation tool such as Playwright or CDP for external pages.
Guide an installing agent through a staged PPT or slide workflow without constraining its implementation choices. Use when the user wants a presentation, deck, storyboard, or slide-ready structure from a topic, article, report, or brief, and success depends on some combination of: clarifying the brief, gathering facts when needed, building an outline, creating page-level planning, generating reviewable intermediate artifacts, and degrading gracefully when the environment lacks research, rendering, or delivery capabilities.
Use when the developer wants the agent to work without approving every write — “tu dong hoan toan”, “cu lam het di”, “dung hoi tung buoc”, “work autonomously”, “approve everything for this task”, a long build session — or when an apply preview says autopilot is active, paused, expired or not covering a statement. Requests one developer-approved lease, works strictly inside it, stops where it stops, and reports every write it made.
Turn what this session learned into durable changes to the repository: a lint rule, a test, a check, a skill edit, AGENTS.md or docs/DECISIONS.md. Use when wrapping up a task, after a mistake or a correction, or when the person says reflect or remember this.
Creates minimal, effective AGENTS.md files using progressive disclosure. Triggers on "create agents.md", "refactor agents.md", "review my agents.md", "claude.md", or questions about agent configuration files. Also triggers proactively when a project is missing AGENTS.md.
Plan and build a new callable Recursive Language Model with predict-rlm. Use when a user wants to design an RLM, assess whether an RLM fits a workflow, or implement an RLM package with structured boundaries, skills, host-side tools, custom adapters, or runtime-kernel extensions. For optimization of an existing RLM, use rlm-gepa instead.
Use Roundfix to plan releases with the read-only Release Plan Command, clean CodeRabbit pull request feedback, diagnose runtime readiness with the Doctor Command, execute a Spec's Task Graph with the Implement Command, monitor Runs through the Supervisor Run Event Stream, reclaim Run storage with the GC Command, archive completed Specs, run its `--follow-up`, `--status` and `--unattended` Skill Modes, and, inside daemon-assigned Batch runs, follow the bounded Review Issue or Task resolution contract.
Build or maintain a Codex-native self-improving memory loop using global `AGENTS.md`, a persistent memories directory, and optional nightly refinement automation. Use when Codex needs to adapt OpenClaw-style self-improvement ideas to Codex, set up long-term user/profile memory, create `PROFILE.md` / `ACTIVE.md` / `LEARNINGS.md` / `ERRORS.md` / `FEATURE_REQUESTS.md`, add promotion rules from raw learnings into active guidance, or create a recurring memory-refinement automation.
Main senior UI/UX orchestration skill. Use this skill whenever the user asks to build, refactor, audit, critique, plan, or validate web app UI, SaaS dashboards, internal tools, landing pages, product pages, screenshots, Figma designs, Figma MCP workflows, design-to-code, code-to-canvas, design-system sync, Stitch/AI UI exploration, three-direction design exploration, cursor reveal effects, responsive behavior, accessibility, visual quality, senior UI/UX/SEO/LLM council decisions, wiki context gates, or UI skill-system workflows. Routes work to webapp-ui-skill, marketing-site-skill, ux-audit-skill, design-critic-skill, ui-ux-pro-max, stitch-design-bridge, senior-figma-orchestrator, cursor-reveal-hero, image-layer-alignment-validator, task-plan-v2-orchestrator, obsidian-wiki-ingest, and shared validation tools while enforcing local-first privacy, one central chair, no fake tools, evidence reporting, and visual verification.
Use ONLY when the user explicitly asks to discover or evaluate a skill from OUTSIDE the curated catalog — e.g., "find a Stripe skill", "is there a skill for X?", "evaluate this candidate before adopting". Meta-skill, complementary to skill-selector: do NOT auto-invoke on routine apm-management, project-init, or catalog-resident requests. Cross-source survey across vetted registries (Anthropic official, claude-skill-registry, VoltAgent/awesome-agent-skills, ComposioHQ, Superpowers, GitHub topic) with a mandatory waxa-eval adoption gate.
Audit claude-skills repository documentation with systematic 9-phase review: standards compliance, official docs verification via Context7/WebFetch, code examples accuracy, cross-file consistency, and version drift detection. Auto-fixes unambiguous issues with severity classification. Use when: investigating skill issues, major package updates detected (e.g., v1.x → v2.x), skill not verified >90 days, before marketplace submission, or troubleshooting outdated API patterns, contradictory examples, broken links, version drift.
Choose the planning form for new work and plan it before execution. Use this skill to decide whether the work is taskless, chat-only, a task file, a task file with subtasks, or multiple task files / backlog items, and to create the needed plan for non-trivial implementation work or investigation work that needs its own task.
A strategic thinking partner for Claude Code that separates deciding from building. Challenges assumptions, compares approaches, and hands execution a ready-to-run prompt in a fresh session. Handles skill routing, context handoff, and memory management. Use when: "plan my project", "advise on architecture", "what should I build next", "help me think through", "explore", "brainstorm", "how should I approach", "what's the right tool", "which skill do I use", "route this task", "hand off context", "manage my session". Triggers on: /strategic-partner, /advisor, /sp
Control Tavus CVI conversations in real-time using the Interactions Protocol. Use when sending text for the replica to speak (echo), interrupting the replica, injecting context mid-conversation, handling tool calls, or listening to conversation events via WebRTC/Daily.
Enable efficient communication between Thai-language users and agents by translating Thai prompts into English in two modes and by preventing Thai text corruption in files. Use when the user writes in Thai, asks for Thai-to-English interpretation, wants token-efficient prompt rewriting, or reports mojibake/replacement-character issues such as U+FFFD in saved files.
Track AI token consumption, costs, and usage trends using the orbit CLI. Use this skill whenever the user asks about token usage, AI costs, Claude Code spending, how many tokens were used, cost breakdown by model, session history, or token analytics. Trigger on phrases like 'how much have I spent', 'token usage', 'show me costs', 'what's my AI spending', 'how many tokens today', 'cost per model', 'list sessions', 'track usage', 'token report', 'weekly usage', 'monthly costs', or any token/cost tracking task — even casual references like 'am I spending too much on Claude', 'what did that session cost', 'show me the dashboard', or 'how much is opus costing us'.
Spin up an isolated git worktree in its own tmux session — coding agent in one window, a shell in another, dev stack in a third — with the agent briefed and already working. Use for spinning up a worktree, starting a feature/branch/task in isolation, handing work to a fresh agent (Claude/Codex), or setting up an isolated tmux dev session.
Move issues and external PRs through a state machine of triage roles, categorise, verify, grill if needed, and write agent-ready briefs. Use when issues or external PRs need categorising, verifying, and turning into agent-ready briefs. Triggers on: triage, triage the backlog, review these issues.
Use when you need to run interactive CLI tools (vim, git rebase -i, Python REPL, etc.) that require real-time input/output - provides tmux-based approach for controlling interactive sessions through detached sessions and send-keys
Implement and deliver an authorized change end to end — verify, open the PR, merge, and confirm post-merge state — using live-state gates, durable interruption handoffs, and live-state-first resumption. Use when a user grants end-to-end delivery authority for a specific change and the work may cross tool-call limits, context exhaustion, worker loss, or session interruptions before the delivery boundary is reached. Do not use for plain fixes with no delivery directive, for granting merge or post-merge authority the user did not grant, or for designing release pipelines (use release-engineering for that).
Write docs/<feature>/HANDOFF.md — the briefing a fresh agent session needs to pick up unfinished work with none of the conversation, covering where the branch and checkout are, what is pushed, what to merge first, what was built and which documents are authoritative, where to resume, and the traps that cost this session time. Use when an implementing run (execute-plan, execute-roadmap, build-feature) stops with work still owed, when context is about to run out, or when the user invokes /write-handoff <feature> or asks to hand work to another agent or session. For handing finished work to a person to test and accept, use write-acceptance instead.
Decompose a spec's PRD/TechSpec into a dependency-ordered task graph — vertical-slice task files plus a machine-parseable _tasks.md DAG manifest under docs/specs/<slug>/, including the authored QA gate decision, decided autonomously from the Spec and escalated only when authority, architecture, or blast radius is at stake.
This skill convenes a 6-seat collaborative pre-production session — Story Architect, Story Analyst, Dialogue Writer, Scene Writer, Standards Reviewer, and Research Specialist pitch their creative visions before writing begins. Optionally convenes the seats across MULTIPLE models/providers (Claude + self-hosted vLLM + cloud) for genuine diversity. Use when: starting a new screenplay, planning major rewrites, reimagining existing material, or seeking diverse expert perspectives.
Evaluate, test, and track an AG2 Agent offline. Build a Suite of tasks, run the agent with run_agent, and grade answers with prebuilt scorers (final_answer_matches, tool_called, no_tool_errors, token_budget) or a custom @scorer — including the agent_judge LLM judge. Read the RunResult scorecard (pass_rate, score_stats, value_counts), gate it in CI with deterministic TestConfig cassettes, persist to store_dir and diff runs to catch regressions, and grade existing traces with evaluate_traces. Use when the user wants to evaluate, test, grade, or benchmark an agent, build a CI or regression gate, or score correctness, tool use, cost, or quality. To compare builds head-to-head or on a leaderboard, see ag2-eval-comparison.
Agent design advisor and review tool. Based on 12-Factor AgentOps best practices, for: (1) discussing Agent architecture designs; (2) reviewing the design of existing Agents/Skills/workflows, finding issues and giving improvement suggestions. Trigger words: review my agent, help me analyze this skill, agent design, agent optimization, help me review this workflow, what's wrong with this agent, how to design an agent, agent architecture consulting.
This skill helps an LLM generate correct core AxAgent code using @ax-llm/ax. Use when the user asks about agent(), child agents, namespaced functions, discovery mode, clarification, bubbleErrors, host-side final/clarification protocol, or ordinary agent runtime behavior. For MCP clients, native runtime modules, subscriptions, tasks, or authentication use ax-mcp alongside this skill. For RLM/code-runtime work use ax-agent-rlm; for callbacks and telemetry use ax-agent-observability; for recall/memory/skill loading use ax-agent-memory-skills; for agent.optimize(...) use ax-agent-optimize.
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.