Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Analyze development sessions, capture learnings, and improve Claude Code instructions. Use when the user wants to reflect on a session, improve CLAUDE.md, extract learnings, or optimize AI-human collaboration. Supports two modes: quick (default) focuses on CLAUDE.md improvements, deep mode performs comprehensive session analysis with learning capture.
Use whenever an agent creates, updates, reviews, or completes a GitHub pull request. Detect whether Robin is installed in the repository; when it is, automatically drive the PR through a bounded review, verified-fix, reply, thread-resolution, re-review, authorized-merge, and cleanup loop without requiring the user to mention Robin. Also use for explicit Robin requests, PR feedback triage, unresolved review threads, or waiting for PR checks.
Integrate the Synthesia Interactive Avatar into a LiveKit Agent end-to-end — detects the user's existing LiveKit setup and attaches the avatar with the three-line pattern, or scaffolds a complete boilerplate voice agent if none exists. Use when: (1) adding a real-time, interactive, conversational, or lip-sync avatar to a LiveKit agent, app, or site, or switching/swapping avatars mid-session, (2) the user mentions the Synthesia avatar plugin, livekit-plugins-synthesia, synthesia.AvatarSession, or the Interactive Avatar, (3) migrating from another avatar provider (HeyGen LiveAvatar, Tavus, Beyond Presence, Hedra) to Synthesia, (4) debugging Synthesia avatar issues — SynthesiaError (type AUTH, UNKNOWN_AVATAR, TIMEOUT, ...), avatar never joins, avatar joins but no lip-sync, (5) building any voice agent where the user wants a human face on it, even if they don't name Synthesia explicitly.
Compose custom agents from base traits, voices, and specializations for specialized perspectives. Create agent teams for parallel work. USE WHEN creating custom agents, spinning up agents, specialized agents, agent personalities, available traits, composing agents, agent profiles, spawning parallel agents, launching agents, agent teams, or swarms.
Use when creating, auditing, or updating AI steering files such as AGENTS.md, CLAUDE.md, GEMINI.md, .github/copilot-instructions.md, .cursor/rules/*.mdc, PRODUCT.md, ANALYTICS.md, TECH.md, STRUCTURE.md, VISION.md, or DESIGN.md in a software repository.
Write an implementation plan that leads with the decisions the user is most likely to change, and buries the mechanical work at the bottom. Use when planning is requested before a build, especially after brainstorming or an interview.
Set up an autonomous research loop where an agent iterates on experiments, hypotheses, and code on git branches. Use this skill when the user wants to automate ML experiments, set up a research agent that self-improves, needs to run many variations of an experiment without manual intervention, or says "automate experiments", "agent research loop", "self-improving agent", "run hyperparameter search", "autoresearch". Based on Karpathy 28k-like AutoResearch post.
Launches a simulated meeting with several expert personas to analyze a topic from diverse perspectives, make a decision, and propose a solution for the user to approve before implementation. Does not write or commit code without explicit user approval; optionally, it can publish the analysis to a linked GitLab or GitHub issue. Use when the user wants to debate an approach, get multi-expert opinions, or receive a recommendation before implementation. For a fully autonomous meeting that implements and delivers directly (commit + MR/PR) without an approval step, use fast-meeting instead
Thin dispatcher / entry point for any Penpot request. Use FIRST on any Penpot task to ensure high_level_overview ran, read the user's intent, and route to exactly one target skill or workflow (build a screen, build a slide deck, build a design system, audit accessibility, audit tokens, migrate from Figma, rename layers, code review, extract DESIGN.md, etc.). Never mutates the canvas. Triggers: 'work on this Penpot file', 'help me with Penpot', 'I want to design/build/audit/migrate in Penpot', 'where do I start', 'which skill should I use', 'route this request', ambiguous Penpot asks.
Implements file-based planning for complex multi-step tasks. Creates task_plan.md, findings.md, and progress.md as persistent working memory. Use when starting tasks requiring >5 tool calls, multi-phase projects, research, or any work where losing track of goals and progress would be costly.
Guided experiment-loop retrospective over the ax agent-experience graph. Walks the user through their open proposals (accept-with-scaffold or reject), pending verdicts (confirm the suggested verdict or override), and recent harness-hook effectiveness signal. Triggers when the user says "let's do an ax retro", "ax retrospective", "review my ax proposals", "triage proposals", "experiment loop status", "lock pending verdicts", "hook effectiveness review", "intervention review", "self-improvement session", or invokes /ax:retro. Reads/writes via the local `ax improve` and `ax hooks` CLIs. Do NOT auto-trigger on unrelated work.
Deep multi-phase implementation planning with parallel agent exploration, iterative user interviews, and structured plan output. Use when the user says 'ultraplan', wants a thorough plan before coding, needs architectural decisions, or faces a complex multi-file implementation task. Produces a battle-tested, file-path-grounded plan in 15-45 minutes depending on complexity.
Discover and use Venice public characters (persona-driven system prompts with a suggested model). Covers GET /characters (search/filter/sort/paginate), /characters/{slug}, /characters/{slug}/reviews, the Character and Review schemas, Bearer-API-key-only auth, filter semantics (adult, pro, modelId), and how to apply a character via venice_parameters.character_slug in chat completions.
Use when the primary task is writing, revising, reviewing, or critiquing agent-assisted coding development text, source-code comments or docstrings, README/docs, CHANGELOG/release notes, PR descriptions, UI copy, chat replies, progress updates, final summaries, or git commit messages, especially when text must be LLM-readable, meaning-preserving, format-bound, language-aware, or evidence-bound. Treat incidental wording inside another active workflow as auxiliary.
Expert delegation specialist that creates comprehensive context packages for coding agents, analyzes requirements, identifies relevant files, and generates clear instructions. Activates when delegating work, assigning tasks, creating delegation packages, or preparing agent instructions.
Create or edit a document that an agent reads, and prove that it changes behavior. Put a lesson that agents keep learning again into structure. Read this skill before you write or edit a skill, AGENTS.md, CLAUDE.md, a system prompt, or other text for agents. Use it to verify such a document before it ships. Use it also when the user corrects the same agent mistake a second time
Transport to the Iskron graph — only when it is unavailable: there are no iskron_* tools or they fail at the transport level (not connected, OAuth, dead token); if even one responds, the skill is not needed. Installs iskron-bridge (stdio bridge, OAuth or personal token) — the path for working; web is a connector with a price. Triggers: “connect the graph”, “raise the bridge”, “iskron tools unavailable”, “mcp not connecting”, “oauth failing”, “log in with a token”, “PAT instead of oauth”, “connect opencode”, “what version is the bridge”, connect to the graph, raise the bridge, MCP tools missing. Composes entry
Launches a fast autonomous meeting with automatically selected personas, then implements the decision, commits, creates an MR/PR, pushes, and publishes a summary in French — all without user intervention. Use when the user wants a decision to be made and executed end to end automatically (including code, commit, and MR/PR). For analysis or a decision without writing code, use meeting instead.
Multi-engine long-term memory for game AI agents. Deep Godot 4.7+ (99 skills, 27 genre blueprints) + Bevy docs-first parity — audited patterns, persona orchestration (Analyst · Auditor · Builder), production-grade engineering.
Shipyard's internal requirements intake for non-engineer contributors (support/ops) — turn a pasted chat log or verbal problem report into a five-section intent.md through numbered batch questioning with a completion gate, walk it through tracker review with harbor's four records, and hand the accepted intent to /launch as its mission brief. Produces decisions, not code.
Add knowledge retrieval (RAG) to AI agents using Twilio Enterprise Knowledge. Covers provisioning Knowledge Bases, uploading sources from web URLs, PDFs, and raw text, and running semantic search to retrieve relevant chunks at runtime. Use this skill to ground agent responses in your organization's actual approved content — FAQs, policies, product docs — rather than hallucinated answers.
Evaluates platform upgrade readiness across Claude model versions, CC releases, and OrchestKit updates with 6-dimensional assessment. Researches target versions, detects current environment, produces structured migration plan with risk scores. Use when planning major version transitions or evaluating upgrade impact.
Help the user pick, write, or apply an AgentRC policy. Policies customise readiness scoring by disabling irrelevant checks, overriding impact/level, setting pass-rate thresholds, or chaining org baselines with team overrides. Use when the user asks about strict mode, AI-only scoring, custom weights, CI gating, or wants org-wide standardisation.
Optimize a complete candidate context package by deduplicating, filtering, ordering, and allocating its token budget. Use when retrieved or assembled material is noisy or exceeds the useful context budget; use context-ranking for scoring chunks and context-compression for shrinking selected content.
Use this skill to set up and configure the Education Cloud Student Recruitment Agent (SRA) — the packaged Agentforce agent (namespace sturecruitment) that answers admissions FAQs, captures inquiries, registers campus tours, and files applications for prospective students. TRIGGER when the user wants to: create or configure an admissions, recruitment, or enrollment agent, set up the Student Recruitment Agent, deploy SRA to an Experience Cloud site, clone the SRA flows or permission sets, add the SRA subagents (Admissions and Enrollments FAQ; Admissions Application; Campus Tours, Visits, and Events Registration; Request for Information), wire Learning Program grounding, or build the escalation subagent. Guides platform enablement, permissions, grounding, agent creation, and channel deployment — API-first with UI fallback. DO NOT TRIGGER for the Transfer Credit Agent, generic Agentforce authoring (use agentforce-generate), or base Education Cloud domain enablement (use education-cloud-domain-configure).
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.