Skills
98,284 packaged capabilities and playbooks for agents. Install any of them with one command.
Development
Write readable, maintainable code through disciplined naming, focused functions, and clean error handling. Use when the user mentions “code review”, “naming conventions”, “function too long”, “code smells”, “readable code”, “boy scout rule”, “single responsibility”, or “unit test quality”. Also trigger when reviewing pull requests for readability, refactoring messy functions, debating comment styles, or improving error handling patterns. Covers SRP, comment discipline, formatting, and unit testing. For refactoring techniques, see refactoring-patterns. For architecture, see clean-architecture
Review code for readability, then simplify it to be easy to follow and understand. Use this skill when the user wants to simplify code, improve readability, clean up for handoff, make code skimmable, reduce complexity, refactor for clarity, strip boilerplate, prepare code for future maintainers, or make a codebase easier to pick up. Trigger liberally — any request touching readability, simplification, cleanup, or "make this easier to understand" should use this skill. Also trigger when the user says things like "review this", "clean this up", "tidy this", "make this simpler", "reduce complexity", "flatten this", "strip the noise", or "I want to understand this code quickly next time I open it".
OpenAI's model connecting vision and language. Enables zero-shot image classification, image-text matching, and cross-modal retrieval. Trained on 400M image-text pairs. Use for image search, content moderation, or vision-language tasks without fine-tuning. Best for general-purpose image understanding.
Create Git commits for the current request's in-scope changes as logical Conventional Commits. Always use this skill for an actual Git commit operation, including requests to commit or amend changes, save work in repository history, stage and record selected files, or group staged or unstaged changes into commits. Use it even when the request describes recording changes without saying "commit"; it performs the commit rather than only suggesting a message, and it does not push.
Based on an already-reached design consensus, research and compare suitable technical libraries, packages, or frameworks, and output an evidence-backed technology selection report with recommendations. Use after consensus is reached in grilling, grill-me, or grill-with-docs when you need technical research, framework comparison, dependency selection, or to prepare technical decisions for to-spec.
End-to-end open source contribution workflow: from scanning issues to submitting PRs. Use this skill whenever the user wants to contribute to an open source project, find issues to fix, submit a pull request, fork a repo to contribute, fix a GitHub issue, or mentions 'open source contribution'. Also trigger when they provide a GitHub repo URL and ask about contributing, say things like 'help me submit a PR', 'find good first issues', 'I want to contribute to X', or mention fixing bugs in someone else's project.
Expert Core Data guidance (iOS/macOS): stack setup, fetch requests and NSFetchedResultsController, saving and merge conflicts, threading and Swift Concurrency, batch operations and persistent history, migrations, performance, and NSPersistentCloudKitContainer/CloudKit sync.
Optimize CoreML models for iOS and macOS deployment. Covers quantization, palettization, pruning, Neural Engine targeting, compute unit selection, and performance profiling. Use when converting ML models to CoreML, optimizing model size/latency, debugging Neural Engine issues, or benchmarking on-device inference.
This skill should be used when the user asks to "show me the changes", "what did you change", "let me see the diff", "review the code", "open web preview of changes", "show me what you worked on", "compare branches", "explain what changed", or wants to view git diffs in a terminal UI, web browser, or get AI-powered code review explanations.
VGV-specific reference for bumping Dart and Flutter SDK constraints across packages, covering pubspec.yaml environment constraints, CI workflow Flutter versions, and SDK upgrade PR preparation. Use it when upgrading the Flutter or Dart SDK version in any VGV repository, on phrases like "bump Flutter to 3.x", "update SDK constraints", "upgrade Dart SDK", "update CI Flutter version", "bump SDK version", or "prep the SDK upgrade PR", and also when a bump is already underway and something breaks: "pub get fails after I changed the environment block", "version solving failed after bumping the SDK", "a dependency requires an older SDK", "get me unblocked on the Flutter upgrade", or "which Dart version ships with Flutter 3.x". It owns the whole bump, including the conflicts it surfaces, even when the blocking package has a skill of its own.
Scripting and automating DaVinci Resolve and building Fusion compositions, macros, and animations. Covers the Resolve scripting API (projects, timelines, media pool, rendering), the Fusion object model, building/editing .setting macros, exposing macro controls, and animating with BezierSpline keyframes, handles, easing, and StyledTextFollower. Use this skill for any DaVinci Resolve Lua/Python script, Fusion macro authoring, or Fusion keyframe/animation work.
Debug web and Node.js applications using console inspection, network analysis, and non-blocking code debugging. Use when the user reports bugs, wants to debug JavaScript (browser or backend), inspect network requests, troubleshoot page/API issues, trace function calls, or monitor exceptions.
Unity of knowledge and action: proactively identify real development work that is starting, in progress, or just finished; run an evidence-based pre-check before the first write to code, Git, or an external system; cite only confirmed constraints during execution; and after completion, once Git/test/MR evidence is obtained, automatically run a retrospective and safe evolution. Development intents in natural language — implementation, modification, debugging, refactoring, fixing, building, development verification, committing code, and Skill feature changes — should implicitly invoke it even if $development-brain is not written; pure product analysis, formal QA, release and deployment, concept explanations, ordinary Q&A, and read-only assessments of whether to update a Skill do not trigger it.
Explore, analyze, plan, or spec features for the Apify MCP server. Adapts to what the user asks — from quick code exploration to full GitHub issue specs. Use when the user asks to explore code, understand behavior, plan a change, design a feature, or create an issue spec.
Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.
Expert Django backend development guidance. Use when creating Django models, views, serializers, or APIs; debugging ORM queries or migrations; optimizing database performance; implementing authentication; writing tests; or working with Django REST Framework. Follows Django best practices and modern patterns.
Django security best practices, authentication, authorization, CSRF protection, SQL injection prevention, XSS prevention, and secure deployment configurations. Use when reviewing Django authentication, authorization, input handling, or deployment settings.
Builds and optimizes language model programs with DSPy (Stanford NLP), using Signatures, modules (Predict, ChainOfThought, ReAct, ProgramOfThought) and optimizers (BootstrapFewShot, MIPRO, BootstrapFinetune). Use when replacing hand-written prompts with declarative signatures, when automatically tuning prompts against training data and a metric, when building multi-stage RAG pipelines or ReAct agents, when building classifiers or structured-output programs, or when saving and evaluating optimized LM programs across providers such as Anthropic, OpenAI, or Ollama. Not for quick prototypes or simple chains; use manual prompting or LangChain instead.
Tracks ML experiments with SwanLab, an open-source tool covering swanlab.init, config and metric logging, scalar charts, and media logging (images, audio, text, GIFs, point clouds, molecules), in cloud, local (mode="local" with swanlab[dashboard]), or self-hosted setups. Includes integrations for PyTorch, Transformers, PyTorch Lightning, and Fastai. Use when logging metrics and hyperparameters for training runs, comparing runs across seeds, checkpoints, or hyperparameters, running an offline or self-hosted dashboard instead of managed SaaS, logging media alongside scalars, or adding tracking callbacks to a Transformers, Lightning, or Fastai training loop. Not for other trackers such as Weights & Biases or MLflow.
Used to extract and generate domain-specific terminology, industry terms, internal and product terms, repository implementation maps, technical stack, and Mermaid architecture diagrams for onboarding from a specified repository, multiple repositories, or a GitHub organization. Triggered when the user requests "create a glossary", "create a domain dictionary", "make onboarding material", "review a repo/org and compile terminology", or "create a knowledge base that the AI doesn't need to re-check".
Evaluate project specifications at inception (Step One), deterministically assess Scale & Load, Domain Decoupling, and Resource Intensity vectors, and automatically enforce a distributed Microservices Architecture when warranted, preventing monolith scaffolding.
Gather the smallest set of repository, environment, documentation, and runtime evidence needed before planning or changing code. Use when load-bearing facts are unknown, tracing execution paths, inspecting unfamiliar packages, or resolving codebase contradictions — even if the user does not explicitly say "fable-discover" (e.g. "explore the codebase", "how does this work", "where is this implemented", "find where this route is handled"). Do NOT use when the bounded edit is already known (use fable-execute), for external API/version research (use fable-research), or for running existing test suites (use fable-verify).
Provision curated capabilities, manage reproducible capability locks, verify host integrations, and compile capability execution contracts for routed get-fable work. Use when discovering machine capabilities, managing eco profiles, planning capability installations, repairing ecosystem health, or preparing capability-aware execution contracts—even if the user doesn't explicitly say "fable-eco" (e.g. "install eco capabilities", "run eco discover", "repair ecosystem", "explain capability run"). Do NOT use as an independent task router (use get-fable).
Implement one accepted, bounded work card with immediate local verification, invariant preservation, and zero scope drift. Use when executing a planned work card, applying a well-defined code change, implementing an isolated function, or performing targeted single-scope edits — even if the user does not explicitly say "fable-execute" (e.g. "implement this card", "write the code for this step", "apply the agreed changes", "build this component"). Do NOT use when new architectural decisions are required (use fable-plan) or when tests are repeatedly failing (use fable-recover).
Automatically synthesize, apply, and verify security remediations for vulnerabilities identified by fable-redteam. Generates surgical AST and code patches, creates continuous TDD regression test guards, performs atomic workspace mutations with automated rollback circuit breakers, and seals cryptographic healing attestations.
Convert discovery evidence into bounded, testable work cards with explicit acceptance criteria and architectural invariants. Use when designing multi-file features, planning complex refactors, decomposing architectural migrations, or structuring multi-step implementation tasks — even if the user does not explicitly say "fable-plan" (e.g. "plan this feature", "design the architecture", "break this task down", "create an implementation plan"). Do NOT use when load-bearing facts remain unknown (use fable-discover first) or for trivial single-scope bug fixes (use fable-tdd or fable-execute).
Diagnose repeated command failures, stale build caches, branch drift, or contradictory evidence before attempting further code edits. Use when commands fail repeatedly, tests stay red after attempted fixes, build output contradicts source code, or the execution path is confused — even if the user does not explicitly say "fable-recover" (e.g. "still failing", "why did this fail again", "stuck in a failure loop", "diagnose this error"). Do NOT use for routine first-time test failures in fresh TDD (use fable-tdd).
Resolve current external facts, official documentation, library behaviors, and API contracts against primary sources before implementation. Use when consulting library documentation, checking breaking changes, investigating external APIs, or verifying framework versions — even if the user does not explicitly say "fable-research" (e.g. "check the latest docs", "what is the API for X in version Y", "look up SDK specs", "verify library support"). Do NOT use for repository-local code exploration (use fable-discover) or for writing implementation code directly (use fable-tdd or fable-execute).
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.