Skills
98,873 packaged capabilities and playbooks for agents. Install any of them with one command.
Development
Install, wire up, and tune glimm—WebGL color-sweep page transitions—in a web project, with correct setup and candid guidance on where it fits. Use whenever the user wants page transitions, route-change animations, more lively or polished navigation, mentions glimm by name, or asks for “view transitions with some color” in a Next.js or React site, even if they never name a library. Also use when customizing an existing glimm install—changing palettes, matching brand colors, adjusting sweep timing or direction, or excluding specific links. Also use when a single in-app moment should get a sweep, such as a plan upgrade, a payment landing, a mode change, or a generated result.
Build from a written issue or task, collaborating with the user: read the issue and related open work, ground it in the code, clarify requirements and design with the user, then implement after they confirm the plan. Use when the user says let's build, build together, or asks to build from an issue or task description.
Use the public `orca` CLI to operate Orca-managed worktrees, folder contexts, terminals, repos, automations, artifacts, 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", 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 browser windows, webviews, or desktop UI outside Orca's embedded browser.
Initialize a new project from flow-next or flow-mono with verified setup, renamed identity, and an independent initial Git commit, while explore shapes the product. Use when the user says "start a Next.js project", "start a new project", or "start a monorepo".
Sync a repository's documentation with what its code actually says: versions, dependencies, commands, project structure and feature claims in README.md, CLAUDE.md, AGENTS.md and docs/, each verified against the manifest, lockfile, scripts and files before it is written, with human prose, tone and section order left as they are. Use when: (1) a dependency, framework or runtime version changed and the docs still name the old one, (2) scripts, Makefile targets or the directory layout changed, (3) a feature was added, renamed or removed and the docs may no longer match, (4) an agent instruction file tells agents to run a command that may not exist, (5) the user mentions: sync docs, update the README, docs are stale, docs out of date, outdated version in README, fix CLAUDE.md, AGENTS.md, package.json, lockfile, pyproject.toml, Cargo.toml, go.mod, "npm run lint" missing script, "docs: sync documentation with codebase state". Carries the ground-truth table per ecosystem, the rule for declared range against resolved version, the flag-don't-guess rule for unverifiable feature claims, and a three-part report of changed, verified and flagged. Works on any project type, Node, Rust, Python, Go or a repository with no manifest at all; the repository's own manifests and files are the seam. Not a changelog or version bump tool, not a doc writer for missing files, and not an audit of the code against the docs
Wire Redis Agent Memory into an existing agent with redis-iris so it remembers users across sessions. Asks five questions (what to remember, what to share, what never to keep, how long to keep it, whether users can be forgotten), writes the answers to service.yaml, provisions the service from the terminal with `iris agent-memory service create`, then wires recall and write-back into the agent's turn in any framework. Use when the agent forgets users, uses InMemoryStore or a checkpointer only, keeps memory in a dict or local DB, or the user asks for long-term memory, recall, or right-to-erasure.
Add Redis Iris Agent Playbooks to an existing agent. Use when approved answers or repeatable procedures live in prompts, document retrieval, or a semantic response cache. Connect or provision a Playbook, publish FAQs and Skills, and verify the first application turn with optional review proposals.
Put the MCP servers an existing agent uses behind one Redis Iris Context Gateway with redis-iris, so the agent connects to one MCP endpoint with one agent key instead of configuring each server and its secret. Asks whether the Gateway is in Redis Cloud or self-managed, then five questions (which MCP servers and secrets the agent has today, how many tools it sees, which read-only tools may be cached, which tools return large results, whether the framework speaks MCP itself). Self-managed, it provisions the Gateway and its Upstreams from the terminal with `iris context-gateway` and sets cache and compression policies. In Redis Cloud the developer creates them in the console. Then it points the agent at the Gateway in any framework and verifies it with `tools list` and `tools call`. Use when the agent configures several MCP servers one by one, carries one secret per server, loads too many tool schemas into the prompt, or the user asks for Context Gateway, one MCP endpoint, tool caching, or result handles.
Replace an existing agent's hand-written routing (keyword if/else chains, regex, or an LLM call that only picks a route) with a Redis Iris Intent Router through redis-iris, so a message maps to a defined intent and its dispatch target by example phrasings and a score threshold. Asks five questions (what the routes are today, a few real phrasings per route, what each route dispatches to, what happens when nothing matches, which embedding model), writes the answers to intents.yaml, provisions the Router from the terminal with `iris intent-router service create`, loads the intents with `iris intent-router intents add`, then wires `router.match()` where the agent decides today. Use when the agent routes with keywords or an extra LLM call, or the user asks for intent detection, intent routing, or semantic routing.
Add Redis Iris Lens tracing to an existing AI agent through redis-iris, so every agent turn is one trace in the Lens UI with its LLM calls, tokens, tool calls, memory and document retrievals, and which of them entered the prompt. Detects how the agent runs (LangGraph or ADK attach, workflows.turn, another framework, or a plain loop), installs redis-iris[llm-tracing], calls lens.init(agent_id=...), opens one lens.turn() per turn where Iris does not open it already, and covers coexistence with an OpenTelemetry setup or LLM instrumentation the app already has. Use when the user asks to add Lens, tracing, observability, or per-turn traces to their agent, or asks why their Lens traces are missing.
Add Redis Iris to an existing AI agent in any framework (LangGraph, LangChain, OpenAI Agents SDK, Google ADK, Pydantic AI, CrewAI, or a plain loop). Use when the user asks to add Iris, Redis Agent Memory, Agent Playbooks, Context Gateway, Intent Router, or Lens to their agent, or asks which Iris product their agent needs. Detects the framework and routes to the product skill.
This skill should be used when the user is building electrobun features for the milady-ai/milady project, submitting PRs to milady that will be reviewed by the agent-review system, asks about "electrobun milady", "milady electrobun PR", trust tiers, Biome compliance, or the release-electrobun.yml workflow.
Researches evidence unavailable locally: external documentation, public repositories at specific refs, archives, and published dependency artifacts (JARs, source JARs, POMs, dependency graphs), reusing a persistent cache of Git clones and Maven artifacts. Use for library docs, upstream source and history, dependency source or metadata, or any question needing external evidence
Use when work arrives at the software factory at any lifecycle stage — a new request, an informational question, a CI failure on a delivered PR, review findings, a merge notification, or a reply to a factory question — and must be routed to the right factory skill, human gate, or follow-up context.
Use when the user asks about MLX serving, mlx_lm.server, oMLX, Apple Silicon LLM serving, or running a local LLM on a Mac, and when troubleshooting a model that fails to load, OOM during load or inference, a server that hangs or crashes at batch>1, tool calls returned as plaintext content, a throughput regression, or the choice between mlx-lm and oMLX. Also covers oMLX feature-flag tuning (turboquant_kv, dflash, MTP, specprefill, thinking_budget, max-concurrent-requests, force_sampling), the OptiQ proxy for models exceeding RAM, Llama-4 ChunkedKVCache batch handling, Llama-3 tool-call JSON format, and bench-driven validation of serving configs. Apple Silicon M-series only, not cloud LLM hosting, non-MLX backends such as llama.cpp, Ollama and vLLM, or model training.
Build NHCX (India's National Health Claims Exchange) payer communication into an existing HMIS: receiving payer queries, notifications and notes on a pre-authorisation or claim, replying with text and documents, and acknowledging notifications, with the embedded NHCX gateway it needs. Use when a hospital system must answer insurer or PMJAY queries over NHCX.
Build NHCX (India's National Health Claims Exchange) payment notices into an existing HMIS: receiving the payer's payment notice and reconciliation breakdown, matching it to the claim, and acknowledging it, with the embedded NHCX gateway it needs. Use when a hospital system must record insurer or PMJAY payments arriving over NHCX.
Build NHCX (India's National Health Claims Exchange) claim reprocess and balance release into an existing HMIS: asking the payer to look again at a decided claim, asking for the unpaid balance of a partly paid claim, and applying the payer's answers, with the embedded NHCX gateway it needs. Use when a hospital system must contest a claim decision or recover a withheld balance over NHCX.
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.
Write the design decision for one change as one interactive HTML page — a zoomable map of the proposed design, the decisions with clickable options, a worked trace, constraints, scope — for the person who locks it. Use when invoked by name (/visual-plan, "make a visual plan", "visual plan for …"), or when a planning or explainer skill asks for its interactive page. Not a task list — the implementation plan is a separate document written after the lock.
Build a website with someone without prior knowledge in eight fixed steps: mockup, design (two or three options based on characteristics), framework, PRD for approval, build and show locally, send to the harbor and show it there in a browser, assembly line, site or server. Open the island in Cockpit and report each step with schritt_melden. Use this skill whenever a website is being built: after webseite-planen, after the inventory in webseite-modernisieren, or for “build me the site”, “mockup”, “draft”, “PRD”, “view locally”, or “what happens after the plan”.
Help a person with no prior knowledge decide whether their website needs a content management system (CMS), and set it up if needed, for example Strapi in Docker on the Albatros server, accessible through a secure tunnel. Explain CMS using the image of a bulletin board. Use this skill for “CMS,” “Strapi,” “change content myself without code,” “blog,” “news,” “manage events,” “my colleague should be able to edit text,” “WordPress,” and “admin area.”
This skill should be used when the user asks to "build electrobun app", "electrobun build", "electrobun canary", "electrobun stable", "set up code signing", "electrobun distribution", or is diagnosing electrobun build failures. Covers dev/canary/stable environments, all three platforms, toolchain prerequisites, artifact output, and CI/CD setup.
This skill should be used when the user asks about cross-platform electrobun support, "electrobun Linux", "electrobun Windows", CEF requirements by platform, "electrobun platform differences", or is setting up multi-platform CI for an electrobun app.
Use this skill when the user asks about “electrobun RPC,” “defineElectrobunRPC,” “Electroview.defineRPC,” “ElectrobunRPCSchema,” bun-to-renderer communication, or renderer-to-bun communication. Activates automatically when RPC code is present.
Use this skill when the user wants to build a complete electrobun feature end to end, asks about the “electrobun SDLC,” “electrobun agent pipeline,” “8-stage electrobun pipeline,” or needs to orchestrate the researcher/architect/planner/dev-squad/QA/test/alignment/docs agents.
Use this skill when the user asks about the “electrobun development workflow,” “electrobun pipeline,” “electrobun lifecycle,” “where am I in the electrobun pipeline,” or needs guidance on what stage comes next in electrobun development.
Use when a Python or TypeScript app sends large tool outputs, logs, JSON, search results or long chat history to an LLM and the task is to cut input tokens, reduce token cost, fit a context window, or trim tool results without losing errors. Adds the headroom-ai compress() call; in-process, no API key, no model download.
Authoring, modifying, reviewing or building openSUSE RPM packages — .spec and .changes files, osc / OBS, and the Git packaging workflow on src.opensuse.org (Gitea). Use for a .osc/ checkout or a *.spec file; when the user mentions osc, OBS, rpmbuild, git-obs, tea, rpmlint or spec-cleaner; or asks to update, build, submit, review or fork a package, check if packages are out of date, or open a package pull request.
Backfill an existing product's screens into your design system's Storybook as representative page templates, rebuilt from the LOCAL design system wherever the product already uses it, with every region that isn't from the system flagged in plain sight. Use when the user says \"get this product into Storybook\", \"backfill our screens\", \"represent this product in the design system\", \"I want to see how a component change ripples across products\", \"relocate these pages into Storybook\", or names a product (a site, an app, a repo, a URL) and a Storybook. Relocation, not translation: it moves screens in and flags what isn't from the system; it never converts bespoke UI onto the system (that's vibe-killer) and never plans a migration (that's ds-adoption-plan). Also runs a refresh that reports how far the product has drifted from its Storybook copy since the last capture.
Connect a Supabase project so its auth emails send through PostHog email templates via the Supabase Send Email Hook, in an order that never breaks a sign-up. Use when asked to set up, connect, test, or fix Supabase auth emails, Supabase email templates, or the Supabase Send Email Hook, or when Supabase shows “email rate limit exceeded” or an auth email error.
Guides the agent through enabling R8 in a Capacitor Android app so that it meets the Google Play DEX code optimization requirement (at least 25% obfuscation, optimization and shrinking from February 2027 for apps with more than 10 MB of DEX code). Covers `minifyEnabled true` with `proguard-android-optimize.txt` in `android/app/build.gradle`, the `-keep @interface com.getcapacitor.** { *; }` rule that prevents the startup crash in R8 full mode, per-plugin keep rules, resolving R8 'Missing class' build errors via `missing_rules.txt`, building a signed release APK with the debug keystore, and verifying with logcat, `mapping.txt` and the `r8.json` file inside the app bundle. Use when Play Console reports a DEX code optimization finding, when a Capacitor app crashes on startup after enabling `minifyEnabled`, or when R8 fails the release build. Do not use for iOS, resource shrinking, Capacitor plugin projects, Capacitor version upgrades, or non-Capacitor mobile frameworks.
Group a scope of Issues into Waves that can be implemented concurrently without stepping on each other — estimate what each Issue touches, exclude the ones that need a human, and print launch lines. Use when the user has several prepared Issues and wants to know which can run in parallel unattended
Explains Git to someone with no prior knowledge using the analogy of a logbook aboard a ship and its copy in the harbor (Albatros's Git service, code.aufwind.sh). Gets the repo from the harbor, guides them through their first entries (Commit) with good notes, sends them to the harbor (Push), and checks that the copy arrives on the Albatros server at ~/projekt. Rules: main only, stage only your own files individually, never force-push. Use this skill for “Git,” “GitHub,” “harbor,” “commit,” “push,” “pull,” “Repo,” “what is a repository,” “save my changes,” “push rejected,” and “conflict,” and before committing anything for the person.
Delete the comments in a diff that the code should make unnecessary, through a fresh read-only reviewer, then fix the code each deleted workaround comment was excusing. Use before a change goes to review, or when the user asks to strip comments from a diff.
Runs shell commands in the background with a fire-and-forget approach and delivers progress notify turns every 2 minutes plus one notify turn to the agent session when each finishes, through the `agent-shell` command (a separate Bun project, installed to `~/.local/bin`). Use when a long command, build, test run or server must keep going without blocking the session, with the result read only after the completion notify turn arrives. Do not use for a quick command that finishes inside one turn.
Sketch types, signatures and module structure before writing code, from several candidate designs on different models, then implement against the chosen sketch. Use when the user asks to architect or design something, or when a skill such as figure-it-out calls for it.
Design and change UI through an interactive shell. Turn a spec and a low-fidelity sketch, or an existing component, into a high-fidelity mock.html in the project's own style, iterate on it with the user, then implement it. Alternatively, open the shell over the running app so the user can inspect, comment on and restyle real elements, with changes going straight into code and tests. Use when the user invokes /building-ui, asks to mock up a new component or UI element from a sketch, change a component through a mock, implement an approved mock, or review or change the UI of the running app.
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.