Skills
101,323 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Map migration-relevant Megatron changes onto the official MindSpeed repository by resolving branch alignment, locating affected subsystems, and identifying concrete adaptation points. Use when Codex has structured Megatron change events and needs to decide whether MindSpeed already covers them, which MindSpeed files are likely affected, and whether patch generation is safe.
Generate multi-style four-grid sticker packs from photos uploaded by the user (built-in Q version, 3D clay, pixel art, and Emoji styles, with support for custom styles), split into 4 separate stickers, and optionally convert them into animated GIFs. Trigger when the user mentions 表情包, 贴纸, sticker pack, sticker, emoji pack, 生成贴纸, 做表情包, Q版贴纸, 大头贴.
Migrate an existing project built for another host (Vercel, a Node server, a Docker container, a docker-compose stack) to Railcode. Use when the user has a working app and wants it running on Railcode, including a Next.js app or a containerized backend in any language. Covers inventorying project dependencies, mapping each dependency to a Railcode primitive, rebuilding components with no direct equivalent another way, porting, testing with railcode dev, keeping or connecting the project's existing database (Postgres or any other) and where schema migrations run, deploying alongside the original, and reporting what Railcode cannot host yet. Do not use for a new app from an idea, or for moving a Railcode v1 app to apps v2 — both belong to create-railcode-app.
Mise (formerly rtx) polyglot tool version manager, environment variable manager, and task runner. Covers tool version management (install, use, pin across Node/Python/Rust/Go/Java and more), mise.toml task definitions (scripts, dependencies, watch mode, incremental builds), environment variable management (.env integration, per-directory env, redaction), project-level tool pinning, backend system (asdf, npm, pipx, cargo, GitHub releases), and comparison with alternatives (asdf, nvm, pyenv, Makefiles). Use when configuring development tool versions, defining project tasks in mise.toml, managing per-project environment variables, setting up polyglot dev environments, migrating from asdf or nvm, or automating build/test/lint/deploy workflows with mise run.
Deploy mkfast-template / TanStarter projects to Cloudflare Workers — automatically detects which components are actually enabled (D1 / R2 / Email / Better Auth / Stripe / Creem / Beehiiv / notifications / analytics) and dynamically trims the deployment steps. Use PROACTIVELY whenever the user mentions deploying, going live, deploying to Cloudflare, wrangler, pnpm deploy, pushing to production, publishing a site, or shipping CF Workers — even if the user does not explicitly say "mkfast", as long as the project contains `wrangler.jsonc` + `@tanstack/react-start` or either `drizzle-orm` / `better-auth`, prefer this skill over the generic wrangler / cloudflare skill (this skill understands mkfast's placeholders, logpush default value traps, enable=false traps, and tailoring strategy). SKIP when the project is clearly not derived from mkfast-template (no wrangler.jsonc and no @tanstack/react-start) → use the generic wrangler skill instead. Triggers: 「部署到 Cloudflare」「deploy to Cloudflare」「mkfast 部署」「TanStarter 上线」「wrangler deploy」「准备部署」「上线前要做什么」「deploy 流程」「pnpm deploy」「推到 Cloudflare 生产」「ship to CF」「go live」「发布 CF Workers」「production deploy」「把网站上线」「上线我的 SaaS」.
Triage a daily msverl regression run by reading the baseline comparison log, stopping on success, extracting the most relevant training failure evidence from the daily training log when needed, collecting recent commits from verl main and MindSpeed master, and ranking the most likely culprit commits with concise fix-direction guidance.
Turn a long video into N viral-ready short clips with a single managed API call. Wraps muapi.ai's `/ai-clipping` endpoint, which handles transcription, highlight ranking through a virality framework (hook / emotional peak / opinion bomb / revelation / conflict / quotable / story peak / practical value), overlap dedupe, and vertical face-tracking auto-crop server-side. No local Whisper, no local LLM, no GPU.
Comprehensive Odoo security auditor for model access rules, HTTP route authentication, sudo() usage, SQL injection risks, and record rule completeness across Odoo 14-19. <example> Context: User wants a full security audit user: "Run a complete security audit on my HR module" assistant: "I will audit access rules, HTTP routes, sudo usage, and SQL injection risks across all files in the module." <commentary>Full audit trigger - comprehensive security review.</commentary> </example> <example> Context: User wants to check access rules user: "Check if all models have proper access rules in ir.model.access.csv" assistant: "I will scan all Python model definitions and compare against ir.model.access.csv to find missing read/write/create/unlink rules." <commentary>Access check trigger - ir.model.access.csv completeness.</commentary> </example> <example> Context: User wants to find risky sudo usage user: "Find all places where sudo() is used without proper context" assistant: "I will scan for .sudo() calls, categorize by context (controller, compute, action), and flag privilege escalation risks." <commentary>Sudo finder trigger - privilege escalation risk analysis.</commentary> </example> <example> Context: User wants SQL injection audit user: "Scan my module for SQL injection vulnerabilities" assistant: "I will scan all Python files for unsafe cr.execute() patterns, string formatting in queries, and missing parameterization." <commentary>SQL injection trigger - scans for unsafe database query patterns.</commentary> </example>
Design structured customer onboarding workflows with phased checklists, email templates, success milestones, and ownership assignments. Use when the user requests onboarding playbook creation or provides relevant inputs for this workflow.
PDF data extraction tool. Use when the user mentions "PDF extraction", "PDF to Markdown", "PDF parsing", "extract PDF content", "PDF to JSON" or "RAG PDF". OpenDataLoader PDF is the current top-ranked PDF parser in benchmarks, supporting local mode (fast, deterministic) and hybrid AI mode (complex tables, scanned documents, formulas), with output in Markdown, JSON (with bounding boxes) and HTML. Suited to scenarios that need structured data extracted from PDFs for RAG/LLM pipelines, or batch processing of PDF documents.
Writes and edits prose in a plain and boring style: simple everyday words, complete sentences, no dashes, no jargon, and no analogies. Use when drafting or revising documents, Notion pages, reports, summaries, READMEs, emails, slides, commit messages, or PR descriptions, or when the user asks to simplify, clean up, tighten, reword, or make writing clearer. Use when the user invokes /plain-writing deslopify or asks to deslopify an agent response. Do not apply to code, only to the words around it.
The CLI for the public API directory at postman.com/explore — search, rank, and watch community-contributed Postman Collections, agent-native and offline. Trigger phrases: `find a postman collection for`, `what postman collection should i fork for`, `is there a postman collection for`, `browse postman api network`, `compare postman publishers`, `what changed on the postman network`, `use postman-explore`, `run postman-explore`.
Apply to any non-trivial work, not just bulk work: edits, migrations, analyses, checks. Build the tool that does it or proves it (codemod, script, generator, or a skill your subagents follow) instead of working by hand. The tool is the artifact a reviewer can rerun
Apply to multi-step work (sweeps, migrations, runs of similar edits) and to how you stack commits and PRs. Break work into small units that each end in a verifiable state, check each before the next, and order delivery so the sequence proves itself to a reviewer.
Comprehensive skill for Product Managers to create organize and manage Product Teams, Strategy, PRDs, release plans, user stories, and standalone issues with high-quality acceptance criteria. Supports both full documentation workflows (PRD → Plan → Stories) and quick issue creation (standalone stories, tasks, bugs). All documentation follows design-driven development principles with functional, testable acceptance criteria based on visual designs and user flows.
Build and run FastFold BoltzGen protein-design workflows end-to-end through API or Composer draft links. Use this whenever users mention BoltzGen, design-spec YAMLs, binder design, multi-spec scaffold workflows, CIF/PDB preparation, workflow graph upsert, `/workflow/composer/<id>`, candidate metrics/structure results, or ask naturally for "help me design a protein" / "give me a simple example".
Tracks quality metrics including defect density, test effectiveness ratio, DORA metrics, and mean time to detection. Use when establishing quality dashboards, defining KPIs, evaluating test suite effectiveness, or reporting quality trends to stakeholders.
Design, automate, and operate end-to-end software releases: release process models and pipelines (trunk-based development, CD stages, release trains), progressive delivery and feature flags, versioning and artifact management (SemVer, conventional commits, changelogs, SBOM/provenance), readiness and quality gates, rollback and recovery planning, change-management and audit compliance (SOC 2, SOX, PCI), DORA metrics, and multi-team release coordination. Do not use for application feature implementation (backend-engineering/frontend-engineering), production incident root-cause debugging or on-call/SLO operations (systematic-debugging / site-reliability-engineering), security implementation or threat modeling (secure-software-engineering), or internal developer platform construction (platform-engineering).
Use when the user wants to automate releases, versioning or changelogs with GitHub Actions; set up or configure release-please (pipelines, pre-releases, monorepos); write release-please-compatible Conventional Commits; tag versions or publish to npm, PyPI, crates.io, Maven or Docker automatically; or troubleshoot why a release PR wasn't created — even without naming release-please ("automate my npm releases", "semver automation", "pre-release workflow"). For multi-repo git operations with submodules, use multi-repo-git-ops instead.
Conducts comprehensive requirements review including completeness validation, clarity assessment, consistency checking, testability evaluation, and standards compliance. Produces detailed review reports with findings, gaps, conflicts, and improvement recommendations. Use when reviewing requirements documents (BRD, SRS, user stories), validating acceptance criteria, assessing requirements quality, identifying gaps and conflicts, or ensuring standards compliance (IEEE 830, INVEST criteria). Trigger when users mention "review requirements", "validate requirements", "check requirements quality", "find requirement issues", or "assess BRD/SRS quality".
Deep retro of retros - investigation pass that surfaces improvements across older retros and the current ax setup. External AI agent drives the reasoning with high thinking enabled. Triggers when user says "deep retro", "investigate ax", "what should I fix in my setup", "retro of retros", or invokes /ax:retro-meta. Use after `ax:retro` if proposals are sparse or you want broader exploration.
Build a source-backed monitoring system for risk signals and catalysts in an A-share company. Use this skill when turning static research into an update process covering orders, financing, governance, policy, operations, and sentiment. All outputs must include accessible source links.
Install and verify Runta tooling for a repository. Use when Codex needs to set up Runta for a user, install the Runta CLI or SDKs, choose the right package manager, check RUNTA_TOKEN, diagnose installation failures, or produce a clear installation report before running Runta workflows.
Design, name, implement, debug, test, and review Rust declarative and procedural macros, including macro_rules matchers and repetition, hygiene, $crate paths, derive, attribute and function-like macros, proc-macro crate naming, syn parsing, quote generation, diagnostics, cargo-expand, doctests, and trybuild. Use when users need compile-time code generation, a Rust DSL, or guidance choosing -derive versus -macros; keep ordinary trait, generic, or handwritten APIs outside macros unless generation has a clear maintenance benefit.
Generate SaaS product launch and software demo video prompts for Seedance 2.0 on Higgsfield. Use whenever the user wants a product launch video, app demo, software walkthrough, feature showcase, startup promo, tech product reveal, or any SaaS/software marketing video. Triggers on SaaS, app, software, product launch, demo, feature, startup, tech, UI, dashboard, landing page video.
Build UIs with Tailwind CSS v4 and shadcn/ui. Covers CSS variables with OKLCH colors, component variants with CVA, responsive design, dark mode, and Tailwind v4.2 features. Supports Radix UI and Base UI primitives, CLI 3.0, and visual styles. Use when building interfaces with Tailwind, styling shadcn/ui components, implementing themes, or working with utility-first CSS. Triggers on tailwind, shadcn, utility classes, CSS variables, OKLCH, component styling, theming, dark mode, radix ui.
Write and improve agent skills (SKILL.md files). Use when creating new skills, refactoring existing ones, debugging why an agent ignores instructions, or improving compliance. Covers prompt structure, TDD for skills, description optimization, and rationalization bulletproofing.
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.
