expo-router Skill: Writing Expo Router Navigation Conventions into the Agent Context
expo-router is a Framework-category Agent Skill (MIT, 1.0.1) published by the Expo team in expo/skills, intended to guide AI systems in correctly using Expo Router: file-based routing in the app directory, Link and Preview/context menus, native stacks via expo-router/stack, modal and formSheet presentations, NativeTabs, and header search. It can be installed in Cursor and similar AI tools with npx skills@latest add expo/skills. Based on the official SKILL.md, this article outlines its positioning, capabilities, installation, and typical code structure, while clarifying its division of responsibilities with expo-native-ui and expo-animation and providing usage considerations.
Read Morehuggingface-spaces: Building and Maintaining AI Demos on Hugging Face Spaces
huggingface-spaces is Hugging Face's official Agent Skill for creating, deploying, and maintaining applications on Spaces. It covers Gradio, Docker, and Static Spaces; ZeroGPU and dedicated hardware; hf spaces commands; README frontmatter; and troubleshooting. The article explains how it works with hf-cli, provides installation paths for Cursor, Claude Code, and similar environments, and includes reproducible commands for searching related demos, creating a Space, and adding frontmatter, plus hardware considerations for free-tier accounts.
Read MoreTransformers.js: Running Hugging Face On-Device Models in the Browser and Node.js
Hugging Face's official transformers-js Agent Skill guides you to run NLP, vision, audio, and multimodal models using Transformers.js in JavaScript/TypeScript, supports browsers and Node.js/Bun/Deno, and covers conventions such as the Pipeline API, WebGPU/WASM, quantization, and dispose. It also introduces npm installation, enabling it with `hf skills add`, and typical code examples.
Read Morehuggingface-llm-trainer: Fine-Tuning LLMs with TRL/Unsloth on Hugging Face Jobs
huggingface-llm-trainer is an Agent Skill in the official Hugging Face Skills repository that guides coding assistants to perform SFT, DPO, GRPO, reward model training, and GGUF conversion on Hugging Face Jobs-hosted GPUs using TRL or Unsloth. The article outlines its positioning, core capabilities, constraints around hf_jobs() and Trackio, the UV script submission workflow, and pre-training checklist items such as Pro/Team account, HF_TOKEN, timeout, and dataset format, while also providing installation paths for Cursor, Claude Code, and Codex.
Read Moregemini-api-dev: Aligning the Gemini Interactions API Development Paradigm with Agent Skills
Google recommends the Interactions API as the recommended interface for new projects, but assistants still tend to generate outdated generateContent calls and deprecated SDKs. The gemini-api-dev skill in google-gemini/gemini-skills—not the 404 gemini-interactions-api path—uses SKILL.md to steer the model toward the correct google-genai/@google/genai version and toward interactions.create for multi-turn and streaming conversations, function calling, and migration checklists. This article explains its positioning, verified capability scope, and installation methods for Cursor, the skills CLI, and Codex, along with minimal Python and TypeScript examples.
Read Moreclickhouse-best-practices: Constraining ClickHouse Reviews and Recommendations with 31 Rules
An Agent Skill maintained by ClickHouse Inc. that requires checking the 31 rules in rules/ and citing rule names before reviewing schemas, queries, and configurations. It covers primary keys and data types, JOINs and skipping indices, batch writes and mutation avoidance, as well as MCP/CLI connectivity, a seven-step schema discovery process, and secure query workflows. It can be installed with npx skills add clickhouse/agent-skills or clickhousectl skills, aligns with the official Best Practices documentation, and is suitable for teams using AI agents for ClickHouse design and SQL review.
Read MoreCloudflare One: Using Agent Skills to Guide Zero-Trust and SASE Deployment
cloudflare-one is an official Cloudflare Agent Skill for designing, configuring, troubleshooting, and reviewing zero-trust and SASE deployments of Cloudflare One (including Access, Gateway, WARP, Tunnel, WAN, DLP, CASB, and more). The skill prescribes a workflow: classify the task, gather account and traffic context, consult current documentation and APIs, and then propose change sets within guardrails that include validation and rollback; migration use cases are paired with cloudflare-one-migrations. The article describes installation methods (Cursor, npx skills, Claude Code, and Codex) and typical conversational usage.
Read Moreagent-platform-rag-engine-management: Manage Google Agent Platform RAG Engine with Agent Skill
Introduces agent-platform-rag-engine-management (v1.0.1) in the Google skills repository: using the vertexai and google-genai SDKs to list RAG corpora and files, view corpus details, retrieve context via retrieval_query, and perform grounded generate_content after user confirmation. It covers installation with npx skills add, gcloud ADC authentication, workflow decision-making, and security tiers Tier R/RC, and clarifies scenarios not suited for SQL, Workspace RAG, or gRAG.
Read Morefirebase-basics: Writing the Firebase CLI Getting Started Workflow into an Agent Skill
firebase-basics from Google's google/skills repository is categorized as Serverless and focuses on Firebase CLI installation and version checks, login (including --no-localhost), project creation and firebase use switching, and retrieving google-services.json and GoogleService-Info.plist via apps:sdkconfig. It explicitly does not cover implementation topics such as Hosting, Firestore, and Auth. The article explains the npx skills add installation method and principles such as confirming the project ID first and prioritizing MCP and official documentation searches, making it well suited for Firebase project initialization in environments like Cursor.
Read Moregcloud Skill: Constraining GCP CLI Operations for AI Agents
The gcloud Skill (v1.0.0) released by Google in the google/skills repository requires agents to validate syntax with leaf-level `gcloud help` before planning or executing `gcloud` commands, prohibiting the use of web searches to guess flags. It enforces minimizing list data, using single commands without pipes, `--quiet` and explicit `--project`/location, along with a four-step plan and dry-run. It covers both direct CLI usage and the MCP `run_gcloud_command`, explicitly listing operations such as IAM, deletion, and billing as prohibited from autonomous execution. Install it with `npx skills add google/skills`.
Read Morecloud-run-basics: Managing Cloud Run Services, Jobs, and Worker Pools with Agent Skills
cloud-run-basics is from google/skills (v1.0.0, Serverless). It uses SKILL.md to teach an Agent to manage Cloud Run Services, Jobs, and Worker pools: enabling APIs, IAM roles, gcloud deployment commands, and constraints such as listening on 0.0.0.0/$PORT and log-based troubleshooting. Install it with `npx skills add google/skills`. It is suitable for teams deploying Agent HTTP backends, scheduled jobs, or pull-based background workloads to GCP, as an operational guide for the official CLI and references, rather than a replacement for security and billing policies.
Read MoreGKE Basics: Using Agent Skill to Avoid Common Pitfalls in GKE Cluster Creation and Workload Deployment
gke-basics is a Containers-category Skill (v1.1.1) in the google/skills repository, designed to guide agents through creating GKE clusters, obtaining kubectl credentials, choosing between Autopilot and Standard modes, and deploying workloads. It also documents common pitfalls, including private clusters and Autopilot CPU request increments. Topics such as Workload Identity, networking, security, and upgrades are routed to dedicated Skills in the same series. It can be installed with `npx skills add google/skills` and used alongside the CLI, MCP, and IaC documentation under references.
Read Morevoltagent-core-reference: VoltAgent Core Class Configuration and Lifecycle Reference
voltagent-core-reference is a reference Agent Skill maintained by the VoltAgent team for the VoltAgent class in @voltagent/core. It documents the constructor options, lifecycle order, and public methods, and marks deprecated fields. It is suitable for coding assistants to reference on demand when configuring MCP/A2A, separating memory between agents and workflows, connecting to a server or serverless runtime, or using VoltOps observability. Install it with npx skills add https://github.com/VoltAgent/skills --skill voltagent-core-reference.
Read Moreneon-postgres-egress-optimizer: Diagnose and Reduce Neon Postgres Data Egress Costs
neon-postgres-egress-optimizer is a dedicated Skill in Neon’s official agent-skills collection, designed to address high egress costs in Serverless Postgres caused by query over-fetching. It follows a four-step process—diagnose, analyze, fix, and verify—and uses pg_stat_statements to identify queries with high row counts, wide columns, and high execution frequencies. It also maps these findings to common code-level anti-patterns, such as SELECT *, missing pagination, application-side aggregation, and redundant JOINs, and provides remediation examples. You can install it by pointing npx skills add to the GitHub repository, and it is recommended to enable the parent neon Skill first. It is well suited for Neon users who want cost-focused query reviews and investigations into billing spikes.
Read Moreclaimable-postgres: One-Click Access to Claimable Postgres for Agents
claimable-postgres is an Agent Skill maintained by Neon that guides coding assistants to provision ephemeral Postgres databases instantly via the Claimable Postgres service. A single API or CLI call returns a real connection string without requiring an account, login, or credit card. It is suitable for adding a DATABASE_URL to new projects, building demo prototypes, running integration tests, and automatically configuring databases during Vite startup. Unclaimed instances expire after 72 hours and can be claimed to a Neon account for long-term use. Installation: run npx skills add specifying neondatabase/agent-skills and claimable-postgres.
Read MoreCallstack GitHub Skill: Standardizing PR and Stacked PR Workflows with the gh CLI
Callstack's GitHub Agent Skill offers standardized workflows for PR creation, stacked PRs, squash merges, rebasing, and repository automation through the gh CLI, targeting teams that rely on PR-based workflows. The official installation command is npx skills add, specifying the agent-skills repository and selecting the github skill. The article introduces its positioning, core capabilities, five typical use cases, and notes on gh authentication and stacked merge order, and provides the officialskills.sh and official GitHub paths.
Read MoreSharp Edges: Finding Security Footguns in APIs Through a Design-Review Lens
sharp-edges is an Agent Skill provided by Trail of Bits for evaluating whether APIs, configurations, and interfaces are resistant to developer misuse. It focuses on six categories of footguns, including algorithm selection, dangerous defaults, configuration cliffs, and silent failures, and follows the principle of pit of success. It provides sharp-edges-analyzer and a four-stage analysis workflow. Install it with: `npx skills add https://github.com/trailofbits/skills --skill sharp-edges`. It is suitable for API/configuration design reviews, but does not replace audits of implementation bugs or business logic.
Read Moremeta-skill-creator: Creating and Validating Agent Skills Per Open Standards
meta-skill-creator is a Meta Skill maintained by oakoss. Based on the agentskills.io open standard, it guides building Skills for compatible agents such as Cursor and Claude Code, covering the Skill directory structure, the five-section format of SKILL.md, frontmatter trigger descriptions, the separation of reference and scripts, and validation and distribution using skills-ref validate. The article explains progressive disclosure, naming constraints, and common errors, making it a useful reference for getting started with the Skill ecosystem and standardizing team practices.
Read MorePrompt Engineering: Writing Agent Prompts as Reusable Engineering Standards
The prompt-engineering skill (v1.0, MIT) in oakoss/agent-skills organizes Chain-of-Thought, ReAct, Tree-of-Thoughts, extended thinking, and tool-call design into installable Agent Skills, with differentiated patterns for Claude 4.x, GPT-4.1, OpenAI o-series models, and multimodal scenarios. The article covers its core principles, pattern quick-reference, common mistakes, and pnpm dlx installation, explains that it suits structured reasoning and Agent workflows rather than simple Q&A or open-ended creative writing, and provides the official GitHub link.
Read Moredsh-window-bounds: Remember the Window Size and Position of the DeepSeek Harness Desktop App
`dsh-window-bounds` is a memory plugin for the DeepSeek Harness desktop app that saves and restores window dimensions and position. Because the Electron shell does not expose the relevant IPC channels, the plugin listens for the `index-inject` event and uses `window.resizeTo` and `window.moveTo` on the `dsh-app://app` page to implement this functionality. It loads data from localStorage, validates it against the current display, and writes the results back to storage with debouncing; on a single display, it restores the position only when it fits. The plugin is enabled by default, has no configuration panel, records only size and position, and does not restore maximized or fullscreen states.
Read Moreagent-patterns: Multi-Agent Design Patterns and Orchestration Practices
agent-patterns is a Multi-Agent-focused skill in the oakoss agent-skills collection (112 hand-curated skills; v1.1, MIT). It catalogs patterns such as fungible swarm, pipeline, hierarchical orchestration, and two-stage review, and explains Claude Code subagent configuration paths, tool allowlists, guidance on parallelism and batch processing, and constraints such as subagents not being nestable. It is suitable as a reference when designing custom .claude/agents, delegating batch tasks, or debugging coordination issues; it can be installed with pnpm dlx skills add oakoss/agent-skills -s agent-patterns.
Read MoreProgrammatic SEO: Scaling SEO Pages with Templates and Data
programmatic-seo is an Agent Skill (v2.1.0) in the coreyhaines31 Marketing Skills series, focused on creating SEO pages at scale using templates and data. The article covers its business assessment process, six core principles, 12 Playbooks (directory, regional, comparison, integrations, etc.), a six-step implementation framework, internal and external linking and indexing strategies, pre-launch quality checks and common mistakes, and explains where to store SKILL.md and how to use it in typical conversations. See the GitHub repository for installation details.
Read MoreAI-SEO: Turning AI Search Optimization (GEO/AEO) into Actionable Workflows
ai-seo, from coreyhaines31/marketingskills (SKILL.md v2.9.0), is designed to help content get cited by ChatGPT, Perplexity, Google AI Overviews, and similar AI-powered sources. The Skill includes reference routing, pre-launch context, a four-step AI visibility audit, the structure/authority/presence pillars, and local references such as agent readiness and llms.txt. It can be installed with `npx skills` or by specifying `ai-seo` using `skillkit`, and is compatible with Agent Skills clients such as Claude Code, Codex, and Cursor. It is well suited for teams that need actionable AEO/GEO checklists. Note that platform behavior and recency data should be considered, strategies differ between Google and multi-engine approaches, and `seo-audit` should be used for traditional SEO audits.
Read Moredsh-task-completion-beep: Play a Browser Notification Sound When a Task Ends
The DSH Web plugin uses the Web Audio API to play five short notification tones when a task's status changes from running to idle, providing clear auditory feedback. It has no external dependencies, is released under the MIT license, and does not access the network or read/write user data. After installation, users can hear the notification sound when a task finishes; if it does not play, they must first interact with the page once to bypass browser autoplay restrictions.
Read MoreCore Web Vitals: Optimizing LCP, INP, and CLS Using Field and Lab Evidence
core-web-vitals is an Agent Skill (v2.0, MIT) in Addy Osmani’s web-quality-skills repository, focused on the three Core Web Vitals: LCP, INP, and CLS. It requires first reviewing CrUX p75 data and browser performance traces, then investigating code and resources only for failing metrics, and retesting with equivalent lab measurements after fixes, without claiming immediate improvements in field metrics. The documentation covers common LCP root causes, checklists, dedicated references for INP/CLS, and the boundaries of measurement sources such as CrUX, Lighthouse, and first-party RUM. Installation: npx skills add specifying the repository and --skill core-web-vitals.
Read More