All agent skills (58,849) - page 126 of 589

Pdf DesignDesign publication-ready PDF layouts for journalism—articles, briefs, and reports—with typography, grids, and export specs agents can follow.402 installs

Qiaomu Opencli AutofixAuto-diagnose and repair common CLI, config, dependency, and script failures encountered during pre-ship review runs and agent-driven fix loops.402 installs

Spatie Securitysecurity-audit402 installs

Spec Driven DevelopmentA three-phase feature-development approach (Requirements, Design, Tasks) that turns vague ideas into well-defined, implementable specs.402 installs

Syncfusion React ChartsUse syncfusion-react-charts for development tasks402 installs

Awq QuantizationQuantize open LLM weights with AWQ (4-bit) and pick GEMM vs GEMV kernels for your inference latency and batch shape.401 installs

AxolotlNavigate Axolotl’s large Python API surface when fine-tuning or customizing LLM training jobs as a ML developer.401 installs

Fine Tuning With TrlAlign an open model on preference pairs with TRL DPOConfig and the right loss variant (sigmoid, IPO, hinge, robust, BCO) instead of guessing hyperparameters.401 installs

Frontend RoutingSet up client-side or hybrid routing, nested layouts, guards, and deep links when scaffolding or extending a React, Vue, or similar SPA.401 installs

Hqq QuantizationConfigure Half-Quadratic Quantization (HQQ) backends and mixed-precision layers so locally run LLMs fit GPU memory without guessing kernel choices.401 installs

Json CanvasAuthor, validate, and transform JSON Canvas files for visual knowledge maps, brainstorm boards, and linked diagram-style documentation in compatible editors.401 installs

LighthouseHelps with ai & agent building tasks.401 installs

Neo4j Document Import SkillTurn PDFs and text corpora into a queryable Neo4j knowledge graph with chunking, extraction, and loader choices your agent can execute end to end.401 installs

Performance MetricsDefine and compute portfolio or business performance metrics—returns, drawdowns, Sharpe, attribution—so dashboards and investor updates reflect consistent, auditable numbers.401 installs

Sentence TransformersChoose a sentence-transformers embedding model that fits your RAG latency, quality, and language needs before you wire up a vector store.401 installs

Threejs Earth RenderHelps with ai & agent building tasks.401 installs

Web Performance AuditUse web-performance-audit for development tasks401 installs

Anti Money LaunderingApply AML controls, suspicious-activity checks, and regulatory reporting patterns when building or operating fintech products, advisory tools, or client onboarding flows.400 installs

AskRoutes prompts to Claude, Codex, or Gemini CLI tools and persists their responses as artifacts within the OMC framework for code review, planning, and UX analysis.400 installs

BubbleteaFor development and infrastructure management.400 installs

Codex ReviewRun structured Codex-assisted reviews on diffs or PRs to catch bugs, style issues, missing tests, and risky changes before merge or release.400 installs

Db Migration HelperHelps with ai & agent building tasks.400 installs

DiversificationResearch and recommend portfolio diversification strategies across asset classes, sectors, and geographies to balance risk and return for client goals.400 installs

Git Hooks SetupHelps with ai & agent building tasks.400 installs

Google WorkspaceRead, create, and update Google Docs, Sheets, Drive files, and Calendar items from an agent workflow for reports, shared specs, and team handoffs.400 installs

Graphic DesignUse graphic-design for development tasks400 installs

LendingSupport consumer or commercial lending workflows covering underwriting criteria, covenant monitoring, and servicing logic in credit products and internal ops tools.400 installs

LlamaguardFilter unsafe user prompts and model outputs with Meta’s LlamaGuard classifier before or after your agent responds in production.400 installs

Openai KnowledgeExtends Claude Code with specialized agent capabilities for developer workflows.400 installs

Openrlhf TrainingPick and configure an OpenRLHF RL algorithm (PPO, GRPO, REINFORCE++, RLOO) with the right flags for your RLHF training run.400 installs

Optimizing Attention FlashPick Flash Attention 2 vs 3 and GPU settings using published forward-pass benchmarks before you ship long-context training or inference.400 installs

OutlinesConfigure Outlines backends—Transformers, vLLM, llama.cpp, or OpenAI—for reliable JSON and schema-bound LLM outputs in agent pipelines.400 installs

Performance ProfilerHelps with ai & agent building tasks.400 installs

PrOpen a Remotion monorepo pull request after format, build, and stylecheck with a pr-name-compliant title and safe gh body-file flow.400 installs

Principle Outcome Oriented ExecutionWhen planning or executing a task, define the user-visible outcome first and trim steps that do not move measurable results closer to done.400 installs

Release Notes GeneratorHelps with productivity & planning tasks.400 installs

Return CalculationsCompute investment return metrics such as CAGR, IRR, and period gains for portfolios, models, and reporting dashboards.400 installs

SentencepieceTrain and choose BPE vs Unigram SentencePiece tokenizers with correct tradeoffs before fine-tuning or serving a custom vocabulary.400 installs

Show Me Your WorkMake Cursor plugin agents expose reasoning, intermediate outputs, and verification steps so users can audit what changed and why during automated edits.400 installs

Sparse Autoencoder TrainingLoad, train, and inspect sparse autoencoders with SAELens for interpretability and feature analysis on language-model activations.400 installs

Tax Loss HarvestingIdentify tax-loss harvesting opportunities, wash-sale constraints, and replacement trades to offset realized gains in taxable accounts.400 installs

Testcontainers Integration TestsRun .NET integration tests with Testcontainers to spin up real databases, queues, and services in disposable containers for faithful pre-release validation.400 installs

Alibabacloud Dataworks Datastudio DevelopDevelop, schedule, and debug DataWorks DataStudio jobs including SQL, Spark, and workflow nodes for analytics and batch data pipelines.399 installs

Code Review SecurityRun AST-based Python security scans before merge or release to catch eval/exec, injection, and secret patterns with severitized JSON output.399 installs

Dart Test FundamentalsWrite and run Dart unit, widget, and integration tests with correct matchers, setup, and CI-friendly patterns before shipping Flutter apps.399 installs

Digital AssetsScope analysis of digital-asset exposure—crypto, tokenized securities, stablecoins—including custody, liquidity, regulatory treatment, and portfolio fit before building or reallocating.399 installs

Expressjs DevelopmentBuild and extend Express.js HTTP APIs and middleware with routing, validation, error handling, and production-ready server structure.399 installs

Fund VehiclesCompare fund vehicle options—ETFs, mutual funds, SMAs, trusts—and scope legal, tax, and operational constraints before product or portfolio design.399 installs

Huggingface TokenizersExplain and apply BPE, WordPiece, and Unigram tokenization so agents train or debug Hugging Face tokenizers correctly.399 installs

Know Your CustomerDesign KYC onboarding, identity verification, and AML screening flows before accepting client funds or opening regulated accounts.399 installs

Long ContextCompare YaRN, ALiBi, and position interpolation when choosing or tuning long-context limits for agent and LLM-backed products.399 installs

Mobile DeveloperImplement native or cross-platform screens, navigation, offline behavior, and store-ready builds while integrating mobile APIs and platform SDKs.399 installs

NanogptBuild and train minimal GPT-2 models to understand transformer architecture and language model internals.399 installs

Nemo GuardrailsWrap production LLM apps with NVIDIA NeMo Guardrails so jailbreaks, toxic output, PII leaks, and hallucinations are blocked at runtime using Colang flows.399 installs

Nginx ExpertConfigure Nginx reverse proxy, TLS termination, upstream routing, caching headers, and load balancing for production web services.399 installs

Peft Fine TuningConfigure DoRA, AdaLoRA, and LoRA+ PEFT adapters when fine-tuning causal LMs without full-weight training on limited GPU memory.399 installs

Principle Never Block On The HumanWhen credentials, approvals, or missing context appear, continue with safe defaults, parallel prep, and documented assumptions instead of stalling the task.399 installs

Publish SkillPackage, document, and publish a finished agent skill to a registry or repo so other Claude Code users can install and run it reliably.399 installs

Readme CreatorDraft complete README files for new repos or packages with install steps, usage examples, badges, and contributor sections agents can refine iteratively.399 installs

Syncfusion React ThemesUse syncfusion-react-themes for development tasks399 installs

Terraform ValidatorReview Terraform plans and HCL for policy violations, unsafe defaults, missing backends, and provider misconfigurations before merge or terraform apply in CI/CD.399 installs

Aem RdeConfigure and work in Adobe Experience Manager Remote Development Environment for local-cloud component development, sync, and preview against AEM instances.398 installs

Ci Cd Pipeline PatternsDesign CI/CD workflows with staged jobs, test gates, build caches, matrix strategies, artifact promotion, and rollback hooks before merging or shipping application releases.398 installs

Convert FileConverts any data file between formats - CSV, Parquet, JSON, Excel, GeoJSON and more - using DuckDB, including binary formats Claude cannot write natively.398 installs

Debugging CodeRun a structured DAP-driven debug loop when hangs, deadlocks, or opaque state block shipping.398 installs

Dotnet Local ToolsConfigure and use dotnet local tools manifests so teams pin CLI utilities—formatters, generators, analyzers—without polluting global SDK installs across repos.398 installs

HigApply Apple Human Interface Guidelines when designing iOS, iPadOS, macOS, and watchOS screens, navigation, and interaction patterns before or during UI buildout.398 installs

Langsmith ObservabilityWire LangSmith evaluate() flows, custom evaluators, and LLM-as-judge scoring so developers can trace and grade agent outputs before and after production.398 installs

Moe TrainingCompare Mixture-of-Experts architectures (Mixtral, DeepSeek-V3, Switch, GLaM) when choosing or explaining sparse LLM designs.398 installs

Nemo CuratorDeduplicate large text corpora before fine-tuning or RAG so developers do not train or index redundant documents.398 installs

Outline CollaboratorCollaborate in Outline wiki: draft docs, organize collections, cross-link pages, and keep team knowledge bases current during product build and ops.398 installs

Persuasion PrinciplesCraft landing copy, emails, and onboarding messages using proven persuasion frameworks to improve signups, conversions, and retention touchpoints.398 installs

Postgresql ExpertDesign PostgreSQL schemas, indexes, migrations, SQL queries, and transactional patterns for relational application backends.398 installs

Segment Anything ModelImplement Meta SAM, SAM 2 video tracking, and Grounded SAM text-to-mask pipelines when you add segmentation to a CV or agent vision feature.398 installs

Sentry Nextjs SdkAdd Sentry error monitoring, performance tracing, and session replay to a Next.js app during implementation before production launch.398 installs

Simpo TrainingPrepare and configure preference datasets in the correct JSON schema before running SimPO fine-tuning on instruction-following models.398 installs

Speculative DecodingUnderstand and apply lookahead (Jacobi) decoding to speed up autoregressive LLM inference without a draft model when you are shipping or tuning agent products.398 installs

Tooluniverse Clinical Trial DesignScope inclusion criteria, endpoints, arms, and feasibility for clinical trials using agent-guided design checks before committing protocol and engineering resources.398 installs

Transformer Lens InterpretabilityLoad transformer models in TransformerLens, hook activations, and inspect weight matrices for mechanistic interpretability experiments.398 installs

Websocket Realtime BuilderUse websocket-realtime-developer for development tasks398 installs

Alicloud Ai Audio Tts Voice CloneIntegrate Alibaba Cloud voice-cloning TTS to reproduce branded or user-specific voices in apps, dubbing pipelines, or personalized agent experiences.397 installs

Angular Best Practices MaterialImplement Angular Material components, theming, accessibility, and layout patterns correctly when building enterprise web UIs with consistent design system constraints.397 installs

Bet SizingSize positions and risk budgets when validating trading, investing, or wagering strategies so capital allocation stays disciplined before live execution.397 installs

Bmad Agent PmUse the BMAD PM agent to refine scope, write epics and stories, sequence work, and keep delivery aligned with product goals.397 installs

Cmux RefHelps with ai & agent building tasks.397 installs

Constitutional AiImplement or study Anthropic-style Constitutional AI and RLAIF workflows when aligning your own models for harmlessness without relying on human harm labels.397 installs

Core DumpsAnalyze native process core dumps after crashes using gdb/lldb, symbol tables, backtraces, and memory/register inspection to isolate segfaults and heap corruption in production.397 installs

Distributed Llm Pretraining TorchtitanConfigure TorchTitan distributed checkpointing, async saves, and seed checkpoints when pretraining or continuing large LLM runs GPUs.397 installs

Economic SystemsAnalyze markets, incentives, and feedback loops when designing products, policies, simulations, or agent scenarios involving money and trade-offs.397 installs

Github Actions ValidatorValidate existing GitHub Actions workflows for syntax errors, insecure patterns, missing permissions, and brittle job dependencies before merge.397 installs

Gitlab Ci GeneratorGenerate .gitlab-ci.yml pipelines with stages for build, test, security scan, and deploy aligned to GitLab runners and project layout.397 installs

GptqPick calibration data and run GPTQ quantization so compressed LLMs stay accurate instead of collapsing perplexity.397 installs

GuidanceConfigure Microsoft Guidance backends—Anthropic, OpenAI, and local models—with correct options when building constrained, template-driven LLM flows in Python.397 installs

Image Video GenAdd image and video generation pipelines to apps or agents, including prompts, model selection, async jobs, storage, moderation, and delivery of generated media assets.397 installs

Lambda Labs Gpu CloudSpin up Lambda Labs GPU instances and wire PyTorch distributed data parallel training across multiple nodes with torchrun.397 installs

Miles Rl TrainingConfigure and run large-scale GRPO/MoE reinforcement-learning training jobs on top of slime with miles-specific FP8, routing replay, and speculative rollout options.397 installs

Openrouter ModelsPick and configure OpenRouter models by cost, latency, context window, and capability when wiring LLM routes into agents, APIs, and production chat features.397 installs

Regulatory ReportingProduce, validate, and file recurring regulatory reports—trade, position, and firm submissions—with reconciliation, attestations, and amendment handling.397 installs

All agent skills - page 126 of 589

This week in AI coding