Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Display Oracle philosophy — the 5 Principles + Rule 6. Use when user asks about principles, "nothing deleted", "patterns over intentions", Oracle philosophy, or needs alignment check. Do NOT trigger for "who are you" (use /who-are-you), "what is oracle" (use /about-oracle), or session status questions.
Comprehensive skill for project planning and prompt engineering. Covers hierarchical plans (briefs, roadmaps, phases), Claude-to-Claude meta-prompts, and multi-stage workflows. Use when: planning, prompt creation, agentic pipeline work, project roadmap, meta-prompts, research to implement workflow.
Prompt engineering and optimization for AI/LLMs. Capabilities: transform unclear prompts, reduce token usage, improve structure, add constraints, optimize for specific models, backward-compatible rewrites. Actions: improve, enhance, optimize, refactor, compress prompts. Keywords: prompt engineering, prompt optimization, token efficiency, LLM prompt, AI prompt, clarity, structure, system prompt, user prompt, few-shot, chain-of-thought, instruction tuning, prompt compression, token reduction, prompt rewrite, semantic preservation. Use when: improving unclear prompts, reducing token consumption, optimizing LLM outputs, restructuring verbose requests, creating system prompts, enhancing prompt clarity.
Mid-conversation reflection skill that pauses execution and zooms out from detail mode to honestly reassess direction, assumptions, and bias. Use when the user says “reflect,” “take a step back,” “step back,” “zoom out,” “are we missing something,” “bigger picture,” “sanity check this,” “are we on track,” “are we overthinking this,” “forest for the trees,” or any variation signaling intent to break out of detail mode and reassess. Also trigger when the conversation has gone deep on implementation details without a strategic check-in, or when the user shows signs of being stuck — that’s often a signal the framing needs a reset, not more detail work. Intentionally low-intake: runs the 5-dimension analysis immediately when prior context is rich enough; asks one forcing clarifier only when invocation context is too thin to reassess from
Use when goal, constraints, and success criteria are already settled and the user wants them frozen into a formal spec an agent can execute and verify: 'turn this into a spec', 'write it up as a Seed', or right after an interview. Returns an immutable Seed with acceptance criteria. Not for vague requests (use interview first).
AstrBot plugin development reference and workflow guide. Use this skill when you are: - Writing AstrBot plugins, hooks, decorators, or message handlers - Implementing platform adapters, message chains, or event flows - Configuring plugin schemas, sessions, or lifecycle management - Working with Agent system (tools, subagents, personas, sandboxes, cron jobs) - Converting between AstrBot message models and platform-specific formats - Looking up AstrBot API signatures, hook inventories, or code entrypoints - Debugging plugin issues related to message routing, event handling, or provider integration Provides reference docs for: messages, platform adapters, plugin config, agent system, and core concepts.
Transforms documents containing theoretical knowledge or frameworks (PDFs, markdown, book notes, research papers, methodology guides) into actionable, reusable Claude Code skills using systematic reading methodology. Use when user mentions "create a skill from this document", "turn this into a skill", "extract a skill from this file", or when analyzing documents with methodologies, frameworks, or processes that could be made actionable.
Create or audit Claude Code skills. Synthesizes source material (articles, frameworks, talks) bottom-up into strategic, interactive skills — or audits existing skills through a multi-lens framework. Use when creating new skills, turning articles into skills, capturing AI workflows as skills, auditing/reviewing existing skills, or improving skill quality. Triggers on: create skill, build skill, new skill, skill from article, turn this into a skill, capture this as a skill, audit skill, review skill, evaluate skill, improve skill.
Use the skill-hub CLI to search, install, update, and publish agent skills from skill-hub.com. Use when you need to fetch new skills on the fly, sync installed skills to latest or a specific version, or publish new/updated skill folders with the npm-installed skill-hub CLI.
Security-first skill vetting protocol for AI agents. Use before installing any skill from the platform skill market, skillhub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns to determine whether a skill is safe to install.
Instructions for AI agents to use the local smart-search CLI for current web search, source-backed fact checking, URL fetching, site mapping, official/API/documentation search, deep research, and reproducible evidence instead of MCP tools or native web search.
Set up a "smart lead + cheap execution" layered sub-agent setup on a new machine or in a new project: a Codex Sol lead + Luna workers, with optional project-level Claude Code agents and Pi/pi-flow cross-tool orchestration. Use for: (1) installing and configuring Codex / Claude Code / Pi from scratch (2) writing project-level .codex/agents, AGENTS.md, .claude/agents (3) fixing the multi-agent catalog problem where Sol cannot spawn Luna (4) running Sol/Luna/multi-agent smoke validation. Triggers: new machine setup, Sol-Luna, layered sub-agents, multi-agent configuration, codex agents initialization
Meta-skill for designing orchestrator+phases structured workflow skills. Creates SKILL.md coordinator with progressive phase loading, TodoWrite patterns, and data flow. Triggers on "design workflow skill", "create workflow skill", "workflow skill designer".
Generate A2UI 0.8 protocol compliant UI code. Use when building agent-driven interfaces, generating JSONL messages for surfaceUpdate/dataModelUpdate/beginRendering, creating forms, lists, cards, or any UI components that render across web/mobile/desktop platforms.
Add agent-friendly --json NDJSON output to Python CLI scripts, or scaffold a complete cli_utils package for a project. Use this skill when the user wants to make scripts machine-readable for AI agents, add --json flags, convert print statements to structured JSON, build a CLI helper library, create an open-source CLI-for-agents package, add structured logging, or make CLI output machine-readable. Also use when the user mentions NDJSON, structured CLI output, agent-friendly CLI, non-interactive scripts, or JSON I/O for automation.
MCP Agent Mail - Mail-like coordination layer for multi-agent workflows. Identities, inbox/outbox, file reservations, contact policies, threaded messaging, pre-commit guard, Human Overseer, static exports, disaster recovery. Git+SQLite backed. Python/FastMCP.
Design and operate AI governance: principles, decision rights, risk tiers, lifecycle gates, fairness, transparency, privacy, security, compliance mapping, maturity and board reporting across SaaS, API, self-hosted and agentic systems. Use to admit, update or retire Agent Skills, MCP servers and A2A capabilities; approve an agent for production actions; or decide earned-autonomy promotion, reduction or revocation. Cover GxP, ALCOA+, data integrity, electronic records, assurance and QMS interfaces for life sciences. Do not use for legal interpretation (legal-strategy), data-platform mechanics (data-architect/data-engineering), security implementation (secure-software-engineering), skill package inspection (agent-skills), or executing runtime changes (agent-production-operations).
Generates a fully structured, project-specific CLAUDE.md by scanning the repo for signal files (package.json, Makefile, .github/workflows/, .env.example, etc.), inferring stack, commands, architecture, and conventions, then asking at most 3 targeted questions for what cannot be inferred. Use whenever the user says 'create a CLAUDE.md', 'my CLAUDE.md is blank', 'generate project context', 'initialize CLAUDE.md', 'set up Claude context for this repo', 'fill in my CLAUDE.md', or any variation where someone needs their project context documented for Claude. If a CLAUDE.md already exists with content, runs a diff-and-merge flow before writing.
Reapply and repair Windows Codex Desktop after Store upgrades, including custom provider models hidden by Statsig available_models filtering, the dependent blue-purple Power slider and its Ultra toggle, Fast Mode request/UI gates, locale i18n, plugin UI gates, Chrome/browser_use gates, Goal command gates, Windows Computer Use availability gates and plugin/runtime repair, phone remote-control pairing under third-party/API-key main app usage, Desktop dynamicTools/inputSchema thread-start schema drift, local conversation visibility recovery after model_provider switches, restored-conversation missing-cwd continuation repair, ASAR integrity repair, signing/installing patched MSIX packages, SDK cleanup, Fast Mode wire verification, local plugin marketplace registration, and optional custom model_instructions_file setup (the bundled system prompt, also called 听话水).
Compress a prompt while preserving semantic content. Use when user says "compress this prompt", "make this shorter", "reduce token count", or "shorten these instructions". Supports lossy (default, 30-50% reduction) and lossless (--lossless, 100% retention) modes.
Build applications powered by GitHub Copilot using the Copilot SDK. Use when creating programmatic integrations with Copilot across Node.js/TypeScript, Python, Go, or .NET. Covers session management, custom tools, streaming, hooks, MCP servers, BYOK providers, session persistence, custom agents, skills, and deployment patterns. Requires GitHub Copilot CLI installed and a GitHub Copilot subscription (unless using BYOK).
Main entry point for the Doubaoya (doubaoya / Bendya) WeChat official account toolbox: onboarding, pre-task routing, post-task navigation, routing users to the right dby-* skill. When the user mentions an official account task but doesn't name a specific skill, or says "which one should I use", "what can Bendya do", "I don't know where to start", route with this skill first instead of guessing a dby-* skill and running it directly. Triggers: /dby, /dby onboarding, Bendya, Doubaoya, "help me see what's next", "what do I do next", "which one first", "where do I start with the official account", "just published one, what now", "just finished writing, then what", "done with this step, what's next", "finished formatting, what else". This package handles routing and hand-offs: **any "just finished a step, what's next" belongs here** (even if what you just finished was writing or publishing); the package itself does not fetch data, write or format. If a specific package has already been named, use that one directly without going through this package. Trigger words: /dby, what's next, where do I start, which dby skill.
Initializes the project for the DO Framework: analyzes it, inventories installed skills and MCP servers (reporting gaps), and writes the framework's local context file `.do/context.md` (summary, conventions, commands, skills and MCPs). Never edits an existing project instructions file (CLAUDE.md, AGENTS.md, .github/copilot-instructions.md, .cursor rules); when the project has none, offers to create one with project content only (no framework, skill or MCP references). Installs the orchestration agents and commands. Arguments: "agents" (install agents/commands only), "refresh" (re-detect skills/MCPs only). Use to initialize or configure the agent-assisted environment. Not for PRDs, implementation, review or QA.
Build DSPy evaluation harnesses with rich-feedback metrics that are essential for GEPA optimization. Use when writing a metric function, calling dspy.Evaluate, splitting dev/val sets, debugging "why is my optimizer not improving?", or designing CI-ready DSPy eval suites.
Effective web research discipline for AI agents — route first, then research with rigor. Step 0 triages whether a question needs external web lookup vs is answerable from the internal codebase/docs; when external, default mode auto-applies 4 maxims; on explicit 'strict/deep research' requests it runs a 7-dimension source-credibility evaluation (CRAAP + E-E-A-T) with an auditable report. When the static layer (WebSearch/WebFetch/curl) can't reach content (login-required / JS-rendered / anti-scraping like 小红书·公众号), it escalates to the real browser via the web-access skill (CDP, carries login state). Triggers: "web research", "external research", "look up information", "effective research", "strict research", "deep research", "strict fact-checking", "logged-in access", "dynamic pages", "anti-scraping", "how to use this library/framework", "are there vulnerabilities", "what is best practice" / web research, look up, investigate, strict research, deep research, fact-check, login-state access, dynamic page, anti-scraping. Do NOT use for: searching the local codebase, reading repo files, grepping code.
Compress natural language memory files (CLAUDE.md, todos, preferences) into simple, plain language to save input tokens. Simplifies jargon-heavy text while preserving all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /eli5-compress <filepath> or "compress memory file"
Entity Memory Management is the systematic extraction, persistence, and retrieval of named entities across agent sessions, enabling long-term contextual awareness that transforms transient conversations into persistent relationships.
Use when a complex problem seems unsolvable and needs digging to the essence — diagnosis only, not fixing. Symptoms include massive logs with no root cause surfaced, complex crashes, architectural essence analysis, requirement authenticity verification, systemic chronic issues, information overload, disconnect between symptoms and essence, multiple intertwined factors, or premature-convergence tendencies — even when the user never explicitly says "diagnose". Also triggers on 本质诊断, 根因诊断, 深度诊断, 诊断问题本质, 梳理问题逻辑, 证据链分析, 逻辑链分析, essence-diagnosis.
Use Chrome DevTools Protocol to allow the AI to "ask Gemini" or "research with Gemini" directly. This uses the user's logged-in Chrome session, bypassing API limits and leveraging the web interface's reasoning capabilities.
Translates vague user prompts into HeyEddi product docs (personas, route intent, voice), route-specific handoff artifacts (wireframes, user mockups, or briefs), and skill-routing under .heyeddi/. Use first on new projects before @heyeddi-design, @heyeddi-handoff, or @flutter-engineering. Never hand-write product.md: use write_product.
Use when adopting incu-way in a repository for the first time — to bootstrap the project so the development, bug, and security flows have what they need. Detects greenfield (new, empty repo) vs brownfield (existing application — the common case), scaffolds CLAUDE.md, the docs/ tree, the branch model, and .ways/, and drives the architectural/functional documentation of the existing code. Also use to re-initialize or refresh a project that already uses incu-way.
Use when a coding agent works in a repository where Jujutsu (jj) is available (a `.jj/` directory) — running jj commands, making speculative edits, splitting changes, recovering with the operation log (jj undo / jj op restore), handing off a PR to Git/GitHub/GitLab, updating a PR after review, or coordinating parallel agents. Uses jj as the local change-management layer while keeping Git as the canonical remote, PR, CI, and audit interface. Also use to avoid jj's agent footguns (pager/editor hangs, commit absorption, colocated git/jj desync) or to replace fragile git reset/checkout/stash/rebase recovery with reversible jj operations.
Best practices for using browser-use's open-source browser-harness with Kernel cloud browsers over CDP. Use when driving a Kernel browser from browser-harness, extracting a CDP URL from `kernel browsers create`, or running multi-step or parallel harness sessions against Kernel.
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.