Browse plugins

A collection of 38 agent skills for practical software engineering including grilling sessions, spec/ticket flows, TDD, code review, and domain modelling—designed to work with any Claude model without external dependencies. These skills emphasize alignment, clear communication, and composability to help engineers avoid common failure modes like misalignment and verbose AI outputs.

This plugin provides behavioral guidelines derived from Andrej Karpathy's observations on LLM coding pitfalls, enforcing four principles—Think Before Coding, Simplicity First, Surgical Changes, and Goal-Driven Execution—to reduce common coding mistakes. It delivers these rules as a reusable skill that constrains Claude Code to avoid overcomplication, hidden confusion, orthogonal edits, and unverifiable success criteria.

Ponytail enforces a "lazy senior dev" philosophy within Claude Code by automatically steering solutions toward the simplest, most pragmatic implementation—preferring stdlib, minimal abstractions, and YAGNI principles. The plugin includes 6 skills that analyze code context, validate security, and catch dependency mismatches to keep output focused, tested, and production-ready.

This plugin scaffolds new Claude Agent SDK applications in Python and TypeScript with a single interactive command, then automatically verifies the setup against best practices using dedicated verifier agents. It handles environment setup, dependency installation, type checking, and provides comprehensive verification reports with actionable recommendations.

This skill automates migration of code and prompts from Sonnet 4.x and Opus 4.1 to Opus 4.5, updating model strings, beta headers, and configuration details. It helps ensure compatibility with the latest model version and can be used iteratively to refine prompts after migration.

This plugin automates code review on pull requests by launching multiple specialized agents in parallel to audit changes from different perspectives, then filters findings using confidence-based scoring to surface only high-confidence issues. It integrates with GitHub, respects repository CLAUDE.md guidelines, and can post reviews as PR comments or output locally.

This plugin provides three slash commands to streamline git workflows: /commit automatically creates commits with AI-drafted messages matching your repo's style, /commit-push-pr handles the complete workflow of committing, pushing, and creating a PR in one step, and /clean_gone removes local branches that have been deleted remotely. The commands analyze your git history and changes to generate contextual commit messages and PR descriptions without requiring external API keys or services.

The feature-dev plugin guides systematic feature development through 7 phases: discovery, codebase exploration, clarifying questions, architecture design, implementation, quality review, and documentation. It orchestrates multiple specialized agents to understand context, design thoughtfully, and ensure quality before and after code is written.

This skill enhances Claude's frontend design capabilities to create production-grade UI/UX with distinctive aesthetics, bold typography, color palettes, and high-impact animations that avoid generic AI-generated interfaces. Claude automatically applies it for frontend work, enabling context-aware implementation of dashboards, landing pages, and other design components with meticulous visual detail.

PR Review Toolkit bundles six specialized agents that analyze different aspects of code quality: comment accuracy, test coverage, error handling, type design, general code compliance, and code simplification. Each agent can be triggered individually with natural language requests or used together for comprehensive pull request review.

UI UX Pro Max is a searchable local database of 79 UI styles, 192 color palettes, 74 font pairings, 25 chart patterns, and 22 implementation stacks (covering React, Vue, Svelte, Flutter, Tailwind, and more) designed to assist with page layouts, components, color schemes, typography, and accessibility. The plugin includes an AI-powered Design System Generator that analyzes project requirements and generates complete, tailored design systems with recommended patterns and styling guidance.

Graphify maps entire code projects (code, docs, PDFs, images, videos) into queryable knowledge graphs using tree-sitter AST parsing for deterministic local code analysis and semantic passes for documentation. Every edge in the graph is tagged as either EXTRACTED (explicitly in source) or INFERRED (resolved by graphify), enabling concept tracing and structured project exploration instead of file grepping.

A comprehensive 25-skill collection encoding production-grade engineering workflows for AI agents across the full software development lifecycle—from specification and planning through testing, code review, and deployment. Skills automatically activate based on context and can be orchestrated via 9 slash commands (/spec, /plan, /build, /test, /review, /ship, and others) or run autonomously with /build auto for end-to-end task execution.

Claude-Mem is a persistent memory compression system that captures tool usage observations and generates semantic summaries across Claude Code sessions, enabling continuity of project context. It automatically compresses and stores session knowledge so Claude can maintain awareness of previous work and decisions even after sessions reconnect.

Taste Skill is a collection of 13 portable Agent Skills that enhance AI-generated frontend interfaces with refined design principles like brutalist, minimalist, and soft aesthetics, covering layout, typography, motion, and spacing. It also includes image-generation skills for creating reference boards that can be paired with AI image generators and handed to code assistants for implementation.

Understand Anything analyzes codebases and knowledge bases with a multi-agent pipeline to build an interactive knowledge graph showing files, functions, classes, and dependencies. The plugin provides a visual dashboard for exploring architecture, searching semantically, analyzing change impacts, and understanding business logic across projects of any size.

ruflo-adr manages the full lifecycle of Architecture Decision Records—creation, indexing, compliance verification, and relationship tracking—by storing ADRs as markdown in a hierarchical AgentDB with causal edges for supersedes/amends/depends-on relationships. It includes git-diff scanning for ADR compliance violations, semantic search, and reconciliation for deleted ADRs.

This plugin provides two agent runtimes for ruflo: local WASM-sandboxed agents (wasm_agent_* / wasm_gallery_* tools) with zero trust and offline capability, plus Anthropic Claude Managed Agents (managed_agent_* tools) for long-running cloud tasks with persistent state. Both runtimes share a unified mental model and API surface, with graceful degradation when keys or dependencies are absent.

Substrate plugin wrapping AgentDB controller bridge (20 MCP tools), RuVector ONNX embeddings (10 tools with RaBitQ quantization), and WASM-backed HNSW pattern router (3 tools) for hierarchical memory, semantic routing, and vector search in Claude Code. Provides safe recall with write guards that keep secrets out of memory, plus health/status monitoring commands and skills for querying and vector operations.

This plugin integrates Ruflo with AGNTCY/Outshift's Internet-of-Cognition runtime, adding SLIM secure transport for cross-host agent coordination, CASA intent-scoped tool authorization, optional IOC coordination events, and observability spans. It is an optional, gracefully-degradable augmentation that remains fully functional even when AGNTCY packages are not installed.

A Claude Code plugin that scans user inputs and system state for prompt injection attacks, jailbreak attempts, and personally identifiable information (PII) like emails, SSNs, and API keys. It provides adaptive threat learning, threat classification with confidence scores, and a three-gate defense pattern for plugins handling untrusted content.

Ruflo-arena is a competitive ruliology system that stages tournaments between deterministic finite-state-machine strategies playing payoff games, producing Wolfram-style competitive arrays and discovering winners through hill-climbing and mutual co-evolution. It implements ADR-147/148 via CLI tools, MCP endpoints, and function-hook mods with full run persistence to `.ruflo/arena/` and AgentDB.

ruflo-autopilot wraps 10 MCP tools to enable autonomous /loop-driven task completion with integrated learning, prediction, and progress tracking across team and swarm tasks. The plugin leverages Claude Code's native /loop and ScheduleWakeup for persistent, cache-aware iteration cycles that discover and apply success patterns from completed work.

This plugin implements cross-host agent federation for AgentBBS via HTTP pull-based synchronization, with Phase 1 room coordination and Phase 2 cryptographic Ed25519 identity and peer pinning to coordinate work ownership across distributed hosts. It includes a security guard tool that prevents credentials and wildcards from being published, and a local `/bbs-mod` command for status and policy auditing.

ruflo-browser provides session-as-skill browser automation using Playwright, capturing each session as a replayable RVF cognitive container with trajectory, screenshots, and findings indexed in AgentDB and protected by AIDefence PII/injection gates. It exposes 10 skills and 23 MCP tools for recording, replaying, extracting, testing, and managing browser sessions with selector memory and cookie vaulting.

A Phase 2 federated business autopilot pod template system that validates sales templates, reserves budget through file-based ledger tracking, and resolves agent roles against a registry before posting envelope summaries via federation. Ships a single-tick runner with dry-run mode by default and implements security guards as function hooks to prevent credential exposure in template paths and contents.

ruflo-daa provides 8 MCP tools for building adaptive agents with cognitive patterns, knowledge sharing, and performance tracking that integrate into the JUDGE phase of an intelligence pipeline. The plugin includes skills for creating agents and defining reasoning patterns, plus a dashboard command for monitoring learning metrics and active agents.

Scaffolds Domain-Driven Design bounded contexts with aggregate roots, value objects, domain events, and anti-corruption layers, storing the domain model as a navigable graph in AgentDB. Validates cross-context boundary violations and detects aggregate invariant issues through CLI commands and an integrated agent.

ruflo-docs automates documentation generation, drift detection, and API docs creation from TypeScript/JSDoc code using the Haiku model for cost efficiency. It drives a background `document` worker that generates full project docs, API references from JSDoc/TSDoc/OpenAPI, and detects when documentation falls out of sync with implementation.

A cross-installation agent federation plugin that enables secure peer discovery, zero-trust mTLS identity verification, and consensus-based task routing with built-in budget circuit breakers (maxHops, maxTokens, maxUsd) to prevent runaway delegation costs. The plugin includes PII detection with adaptive policies (BLOCK/REDACT/HASH/PASS), Byzantine fault-tolerant state mutations, and compliance-grade audit trails for HIPAA, SOC2, and GDPR.

A Claude Code plugin for long-horizon goal planning and research orchestration using GOAP algorithms, featuring deep multi-source research campaigns, adaptive replanning, and persistent objective tracking across sessions. It bundles four specialized agents (goal-planner, deep-researcher, horizon-tracker, dossier-investigator) with skills for action planning, evidence-graded synthesis, and recursive parallel investigation building graph-structured dossiers.

RuFlo Graph Intelligence Engine provides real-time relationship intelligence using single-entry personalized PageRank with streaming delta updates and witness-signed reasoning artifacts for federation distribution. The mod includes a write guard to protect sensitive data (keys, tokens, passwords) in graph operations and provides local status/diagnostic commands without network calls.

ruflo-intelligence is the user-facing surface for Ruflo's self-learning system, wrapping 29 MCP tools into discoverable skills and operationalizing the 4-step pipeline (RETRIEVE → JUDGE → DISTILL → CONSOLIDATE) for continuous pattern learning and consolidation. It coordinates with ruflo-agentdb (namespace conventions), ruflo-ruvector (trajectory recording), and ruflo-browser (session replay) to enable cross-project pattern transfer via IPFS and prevent catastrophic forgetting through EWC++ consolidation.

This plugin manages Cognitum Seed edge devices through four Claude agents (device-coordinator, telemetry-analyzer, fleet-manager, witness-auditor) that handle device lifecycle, anomaly detection via Z-score analysis, firmware rollouts, and witness chain verification across fleets. It provides CLI commands and skills for device registration, telemetry ingestion, fleet CRUD, firmware deployment with canary/rolling state machine, mesh coordination, and trust scoring based on pairing integrity, firmware currency, uptime, witness integrity, and anomaly history.

ruflo-jujutsu provides advanced git workflows with diff analysis, risk scoring, change classification (feature/bugfix/refactor/etc.), and automated reviewer recommendations through 6 MCP tools. It includes PR text and merge/close guards as a function-hook mod to prevent accidental exposure of secrets in pull requests.

Extracts entities (classes, functions, modules, types, concepts) and relations from source code and documentation, then builds a navigable knowledge graph stored in AgentDB with hierarchical nodes and causal edges. Traverses the graph using a pathfinder algorithm that scores paths by edge weight and semantic similarity, with a security mod that prevents secrets from being stored in the graph and requires confirmation for deletions.

This plugin manages cache-aware background workers and cron jobs via 5 MCP tools (list/dispatch/status/detect/cancel), exposing 12 worker triggers for tasks like ultralearn, optimize, audit, and testgaps across the ruflo ecosystem. It includes a security guard mod that blocks secrets in worker contexts and enforces per-session dispatch caps, with local status tracking via `/loop-mod` commands.

Ingests and normalizes market data from REST APIs and WebSocket feeds into OHLCV vectors, then vectorizes and detects candlestick patterns (doji, engulfing, head & shoulders, etc.) for read-only similarity search and pattern matching. A mod guard (ADR-445) prevents credentials and feed URLs from entering market memory namespaces.

ruflo-metaharness surfaces MetaHarness CLI tools (score, genome, threat modeling, security scanning, harness scaffolding, and evolutionary mutation) through 13 ruflo skills, maintaining optional pluggability so ruflo remains fully operational if MetaHarness packages are removed. All skills degrade gracefully on missing dependencies, use subprocess invocation with structured JSON output, and enforce security controls via a mod function-hook that blocks secrets in inputs and gates privileged operations.

This Claude Code plugin generates, validates, and manages database schema migrations with sequential numbering, up/down SQL pairs, and rollback safety checks. It includes dry-run mode, validation checks for foreign key consistency and naming conventions, full migration history tracking, and a security guard mod that prevents writes containing secrets or database URLs with inline passwords.

The ruflo-music plugin wraps the Cognitum Music MCP server to compose lyrics, generate tracks, separate stems, extract MIDI, and master audio through a remote GPU inference service authenticated with a personal access token. It exposes 2 agents and 7 skills for the full music production pipeline, with production metadata cached in AgentDB for fast recall.

ruflo-neural-trader wraps the neural-trader npm package (v2.8.11+) as a Ruflo plugin, providing 4 specialized agents and 9 skills for LSTM/Transformer/N-BEATS neural trading strategies with Rust/NAPI backtesting, Monte Carlo simulation, and portfolio optimization. Heavy computational jobs can be dispatched to Anthropic Managed Agents cloud containers, with optional local execution of backtests, signal generation, regime detection, and risk assessment using Yahoo Finance data.

ruflo-observability implements OpenTelemetry-compatible structured logging, distributed tracing, and metrics collection to correlate agent swarm activity with application-level telemetry, including anomaly detection in latency, error rates, and resource usage. As a function-hook mod, it also enforces a tighten-only guard that prevents telemetry writes containing secrets or sensitive data like SSNs and payment card numbers.

Scaffolds, validates, and publishes new Claude Code plugins with a canonical contract including ADR documentation, smoke tests, MCP tool wiring, and drift warnings for known tool/namespace misconfigurations. Includes a function-hook mod for local plugin validation and reserved-name checking without model calls.

Retrieval-Augmented Generation memory system that uses HNSW vector search and AgentDB to provide semantic store/search/recall across Claude Code sessions, with a write guard (ADR-445) that prevents secrets from entering memory. Bridges native Claude Code auto-memory into persistent AgentDB with 384-dim ONNX embeddings for unified cross-session semantic retrieval.

Wraps ruvector@0.2.25, a Rust-backed vector database with HNSW indexing, adaptive LoRA embeddings, code-graph clustering, and hooks-based intelligence, providing sub-millisecond semantic search and 52,000+ inserts/sec. Includes the vector-engineer agent and skills for embedding, clustering, hyperbolic projection, plus 91 MCP tools for RAG, brain knowledge, SONA learning patterns, security scanning, and diff analysis.

ruflo-ruvllm provides local LLM inference with chat formatting, MicroLoRA fine-tuning, and SONA real-time neural adaptation for models like Llama3, Mistral, and Gemma. It functions as a mod (ADR-445) that guards against secrets being trained into adapters and routes context for RAG pipelines using HNSW with up to 11 patterns.

ruflo-rvf provides RVF format for portable agent memory, session persistence, and cross-platform transfer of Claude Code sessions and agent knowledge. It includes a security mod (ADR-445) that guards against secrets leaking into exported memories, sessions, or transfers, with optional AES-256-GCM encryption at rest.

ruflo-security-audit performs deep security scanning, CVE monitoring, dependency analysis, and policy enforcement for CI/CD pipelines with pattern-based detection of shell injection, plaintext secrets, and environment variable hijacking. It integrates with ruflo-core and complements runtime security gates via static analysis of code vulnerabilities and supply-chain risks.

ruflo-sparc orchestrates a five-phase SPARC development lifecycle (Specification, Pseudocode, Architecture, Refinement, Completion) with enforced quality gates and specialized agents for each phase. It spawns phase-specific agents, tracks all artifacts and gate results in memory, and coordinates with sibling plugins for deeper work at each phase.

ruflo-swarm coordinates multi-agent teams with 12 MCP tools (4 swarm_* and 8 agent_*) and 6 topologies (hierarchical, mesh, ring, star, adaptive, hierarchical-mesh), enabling real-time monitoring and git worktree isolation per agent. With Claude Code function hooks enabled, a live swarm pane displays agent tiles, topology, task claims, consensus votes, cost tracking, and CLI-backed buttons for swarm coordination.

This plugin detects test coverage gaps, analyzes coverage across codebases, and generates automated test scaffolding to improve test coverage with London School (mock-first) patterns. It integrates with the SPARC Refinement phase to enforce ≥80% coverage on new code and coordinates background test gap analysis via the testgaps worker.

ruflo-workflows provides workflow automation across two complementary surfaces: MCP workflow_* tools for declarative, stateful, long-lived pipelines with human approval gates, and native Claude Code Workflow JS scripts for deterministic fan-out orchestration of subagents. It includes state-machine lifecycle management (created → running ↔ paused → completed/cancelled) and a GAIA benchmark component for leaderboard submissions.

headroom-snip visualizes token compression in real time as Headroom optimizes each Claude Code request, displaying an animated band showing original vs. sent token counts and running compression totals. It provides a /headroom command for detailed per-request breakdown and works only with local proxy instances, never sending data remotely.

Impeccable provides design fluency for AI-assisted frontend development through 23 specialized commands and 59 deterministic anti-pattern detection rules. It guides code agents on layout, typography, color, animation, accessibility, and UI polish while enforcing a design system context initialized once per project.

An AI agent-led search engine that aggregates and scores results from Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and 5+ other sources by real engagement metrics (upvotes, likes, real-money odds) rather than editorial curation. The skill searches all platforms in parallel and synthesizes findings into a brief based on what real people actually engage with.

HyperFrames is an open-source framework that converts HTML, CSS, media, and seekable animations into deterministic MP4 videos, available via CLI, AI coding agent skills, or hosted authoring workflows. The plugin provides 21 skills for agents to plan, write, animate, and render videos with support for compositions, GSAP animations, captions, voiceovers, audio-reactive visuals, and website capture.

Humanizer rewrites AI-generated text to sound natural and human-written by detecting and removing 26 common patterns of artificial language like "not X but Y," staged run-ups, and false drama. It preserves factual accuracy and can match the writer's voice by analyzing writing samples.

A collection of 49 AI agent skills covering marketing tasks including conversion optimization, copywriting, SEO, paid ads, analytics, and growth engineering. The skills work together as a modular system, with product-marketing as the foundation that other skills reference for product context before executing specialized marketing workflows.

This plugin provides a curated collection of 2,055 pre-built skills designed for Claude Code, enabling developers to leverage a comprehensive library of ready-made functionality without external dependencies. The skills cover a wide range of use cases and integrate directly into Claude Code workflows.

The AAS Accessibility & Inclusive UX bundle packages 8 skills for Claude Code that help developers build more accessible and inclusive user experiences. The plugin provides guidance and tools for implementing best practices in accessibility standards and inclusive design.

This bundle packages 10 skills for building agentic systems and Model Context Protocol (MCP) servers within Claude Code. It provides editorial resources and templates from Agentic Awesome Skills to jumpstart agent and MCP development.

This Claude Code plugin bundle aggregates 10 editorial skills for AI product evaluation and operations workflows. It provides a curated collection of capabilities from the Agentic Awesome Skills library focused on assessment and operational excellence in AI product development.

This bundle for Claude Code packages 10 skills designed to help build and develop API platforms using AAS (Agentic Awesome Skills). It provides a curated collection of tools for API platform construction without requiring external API keys or dependencies.

The AAS Automation Builder is a Claude Code plugin bundle containing 10 skills for building and automating workflows. It provides a collection of reusable skill modules without requiring external API keys or third-party tool integrations.

This plugin provides a collection of 10 data analytics skills bundled as an editorial "AAS Data Analytics" suite for Claude Code integration. It enables data analysis workflows without requiring external API keys or tools.

This bundle packages 10 data-engineering skills for Claude Code, enabling streamlined workflows for data pipeline development and engineering tasks. No external API keys or third-party services are required.

This plugin bundles 10 DevOps and cloud-related skills from Agentic Awesome Skills for Claude Code. It provides editorial curated functionality for cloud infrastructure and DevOps workflows without requiring external API keys or tools.

This Claude Code plugin bundle provides 9 editorial skills for creating, managing, and presenting AAS (Agentic Awesome Skills) documents and presentations. It requires no external API keys or tools, operating entirely within Claude's native capabilities.

This bundle collects 10 skills for localizing content and supporting international growth strategies within Claude Code. It provides reusable components for language adaptation, market analysis, and global expansion workflows without requiring external API keys or services.

This Claude Code plugin bundles 10 marketing, SEO, and growth skills for editorial assistance and content optimization. It integrates directly into Claude Code without requiring external API keys or tools.

This Claude Code plugin bundles 10 agent skills for building mobile applications, providing a collection of tools to streamline app development workflows. It requires no external API keys or services, making it a self-contained solution for mobile app builders.

This Claude Code plugin is an editorial bundle from Agentic Awesome Skills containing 10 observability and incident response skills for structured analysis and monitoring workflows. It requires no external APIs or tools and runs entirely within Claude Code.

This Claude Code plugin bundles five privacy and compliance engineering skills from Agentic Awesome Skills to help developers implement privacy-preserving architectures and meet regulatory requirements. It provides tools for privacy assessment, compliance checking, and engineering practices without requiring external API keys or services.

This Claude Code plugin bundles 10 skills for product design workflows, enabling designers to streamline design processes and decision-making within their development environment. The bundle operates entirely within Claude Code with no external API dependencies or tool requirements.

This Claude Code plugin bundles 10 skills for building Python APIs with a focus on automated scaffolding and developer productivity. It provides reusable patterns and code generation for common API construction tasks.

An editorial bundle of 10 reusable skills from Agentic Awesome Skills for quality assurance and test automation workflows in Claude Code. The bundle provides ready-made capabilities for QA testing scenarios without requiring external API keys or tools.

This Claude Code plugin provides 10 editorial skills bundled for launching and scaling AAS (App-as-a-Service) and SaaS applications. The bundle equips developers with guidance on product launch strategy, revenue modeling, and growth optimization without requiring external API keys or services.

This bundle packages 9 skills for Claude Code that enable secure application development and building practices. It provides a curated collection of tools focused on security-aware app development workflows.

This bundle contains 10 security-focused skills curated for Claude Code to help developers build, review, and audit software with security best practices in mind. No external dependencies or API keys are required—all capabilities are self-contained within the bundle.

This Claude Code plugin bundles ten skills from Agentic Awesome Skills for building web applications, integrating directly into Claude's code editor without requiring external API keys or tools. It provides a collection of editorial resources and utilities to streamline web app development workflows.

This Claude Code plugin bundles seven Agent Architect skills designed to help developers design, plan, and structure agentic systems within the Claude Code editor. The skills provide guidance and architectural patterns for building multi-agent workflows without requiring external API keys or tools.

This plugin bundles six design skills for Apple platform development, providing editorial guidance and best practices for creating native iOS, macOS, tvOS, and watchOS applications. It requires no external API keys or third-party services, making it immediately usable within Claude Code.

This bundle delivers 5 pre-built skills for architecture and design workflows within Claude Code, curated from the Agentic Awesome Skills collection. No external API keys or third-party services are required.

This Claude Code plugin provides an editorial bundle of 8 automation-building skills from Agentic Awesome Skills, designed to streamline workflow automation tasks. It requires no external APIs or tools and runs entirely within Claude Code's environment.

This plugin provides six skills bundled for cloud and AI development on Azure, offering integrated capabilities for developers working with Azure services and cloud infrastructure. The bundle packages editorial selections from Agentic Awesome Skills without requiring external API keys or third-party service integrations.

The agentic-bundle-business-analyst is an editorial bundle containing 5 reusable skills for Claude Code that support business analysis workflows. These skills provide pre-built capabilities to assist with analytical tasks without requiring external API keys or integrations.

This editorial bundle packages 6 skills for commerce and payments functionality within Claude Code, providing pre-built components for transaction and e-commerce workflows. No external API keys or tools are required; all capabilities are self-contained within the bundle.

This Claude Code plugin bundles six editorial skills designed to assist with creative direction and content strategy. It integrates directly into Claude Code's workflow to support editorial planning and creative decision-making.

This Claude Code plugin bundles seven pre-built data analytics skills for streamlined analytics workflows. The editorial collection from Agentic Awesome Skills provides ready-to-use tools without requiring external API keys or third-party integrations.

This Claude Code plugin bundles five data engineering skills from Agentic Awesome Skills, providing pre-built agent capabilities for data pipeline, transformation, and processing workflows. The bundle requires no external APIs or tools, running entirely within Claude's environment.

This editorial bundle bundles 8 curated skills focused on Domain-Driven Design and Evented Architecture patterns for Claude Code. It requires no external tools or API keys and runs entirely within Claude Code's environment.

This is an editorial bundle of 8 DevOps and Cloud skills for Claude Code that provides curated tools and workflows for cloud infrastructure management and deployment. The bundle packages reusable skill components with no external API dependencies or third-party service integrations required.

This plugin bundles five essential skills for Claude Code, curated from Agentic Awesome Skills to enhance core development workflows. The bundle integrates seamlessly without requiring external API keys or third-party services.

Creates editorial-quality diagrams including architecture, flowcharts, sequences, state machines, ER models, timelines, swimlanes, quadrants, radars, loops, trees, org charts, Venn diagrams, pyramids, Gantt charts, and Wardley maps, outputting self-contained HTML+SVG files with minimal styling and optional motion. The plugin converts existing draw.io, Mermaid, or Excalidraw diagrams to chosen formats and detail levels while maintaining semantic pattern separation from layout.

This plugin audits web interfaces for WCAG accessibility compliance, validating screen reader compatibility, keyboard navigation, and inclusive design patterns. It provides both automated testing commands and agent-driven compliance analysis without requiring external dependencies.

This plugin provides multi-agent system optimization with two commands and one agent for managing agent improvement workflows and context management. It enables teams to coordinate and enhance agent performance within Claude Code environments.

The agent-teams plugin orchestrates multi-agent teams within Claude Code to enable parallel code review, hypothesis-driven debugging, and coordinated feature development. It provides preset team compositions (review, debug, feature, research, security, migration) and structured agent communication protocols for complex collaborative tasks.

This plugin helps developers design and scaffold REST and GraphQL APIs by selecting appropriate frameworks, defining backend architecture, and generating boilerplate code. It includes four agents that guide framework selection, architecture planning, and API generation without requiring external dependencies.

This Claude Code plugin automates API testing, mocking HTTP requests, and generating OpenAPI documentation to validate your API contracts and behavior. It integrates observability setup and monitoring capabilities to help track API performance and issues in development and testing environments.

This Claude Code plugin provides application profiling, performance optimization, and observability capabilities for both frontend and backend systems through integrated agents and commands. It enables developers to analyze and improve application performance without requiring external dependencies or third-party services.

This plugin provides an agent for ARM Cortex-M firmware development targeting popular microcontroller platforms including Teensy, STM32, nRF52, and SAMD with built-in peripheral drivers and memory safety patterns. It enables developers to write and optimize embedded firmware with safety guarantees across multiple ARM-based architectures.

This Claude Code skill audits and rewrites prose that reads as machine-generated across markdown files like READMEs, changelogs, PR descriptions, and blog copy, offering detect-only, rewrite, and edit-in-place modes. It uses a pattern catalog covering formatting, sentence structure, and machine fingerprints, with tiered vocabulary tables and context/voice profiles, while explicitly avoiding false authorship classification and imposing house styles.

This plugin provides two agents for implementing API security practices including authentication, authorization, rate limiting, and input validation patterns. It helps developers harden backend APIs through guided security configuration and best-practice recommendations.

This Claude Code plugin provides 8 agents and 9 skills for backend API design, GraphQL architecture, workflow orchestration with Temporal, and test-driven development practices. It includes a command interface and requires no external API keys or third-party services.

Before You Build is a pre-mortem risk assessment skill that helps founders and product teams identify potential pitfalls and product risks before starting implementation on landing pages, MVPs, SaaS products, or major features. It provides a structured markdown-based review to validate product ideas and catch issues early, avoiding wasted engineering effort on flawed concepts.

This plugin provides 1 agent and 4 skills for smart contract development, covering Solidity programming, DeFi protocol implementation, NFT platforms, and Web3 application architecture. It operates with no external dependencies, API keys, or third-party service integrations.

This Claude Code plugin enforces git hook integrity by preventing AI agents from using bypass flags like --no-verify and --no-gpg-sign during git operations. It implements a PreToolUse hook that blocks these security-circumventing flags at the agent level.

A Claude Code skill that guides developers through structured brand discovery interviews and iteratively generates landing page designs via Stitch, translating brand preferences into deployment-ready HTML. It handles design decisions across color palette, typography, and layout while capturing user feedback to refine the output through multiple variants.

This Claude Code plugin provides business metrics analysis, KPI tracking, and financial reporting capabilities through an agent with two supporting skills. It enables data-driven decision making without requiring external API keys or third-party service integrations.

This plugin automates C4 architecture documentation by analyzing code from the bottom up to identify components, map containers, and generate context diagrams. It orchestrates multiple agents to synthesize architectural insights and produce comprehensive architecture diagrams and documentation.

This Claude Code plugin provides agents and skills for configuring CI/CD pipelines, setting up GitHub Actions and GitLab CI workflows, and orchestrating automated deployment pipelines. It enables developers to define, test, and deploy infrastructure and application changes through automated workflow configuration.

This plugin provides cloud architecture design and infrastructure management across AWS, Azure, GCP, and OCI, with specialized support for Kubernetes configuration and Terraform infrastructure-as-code. It includes multi-cloud cost optimization and hybrid cloud networking capabilities through 7 agents and 8 skills.

This plugin automates technical debt reduction, dependency updates, and code refactoring through 3 commands and 2 agents that analyze and improve codebases. It operates without external service dependencies, making it suitable for self-hosted environments.

This plugin generates documentation and explanations for codebases through automated doc generation and tutorial creation, with 2 commands and 3 agents to handle different documentation tasks. It requires no external APIs or tools and operates entirely within Claude's capabilities.

This Claude Code plugin automates code cleanup and refactoring tasks while preserving application context, helping teams reduce technical debt through intelligent code transformation. It includes 3 commands and 2 agents that work together to analyze and improve code quality.

This Claude Code plugin performs multi-perspective code analysis across architecture, security, and best practices using 2 commands and 3 agents to identify issues and improvements. It operates entirely within Claude's environment without requiring external API keys or tools.

Conductor transforms Claude Code into a context-driven project management tool by enforcing a structured workflow of Context → Spec & Plan → Implement across feature tracks. It manages persistent project artifacts including product vision, technical decisions, and phased work units with git-aware reversion capabilities.

This Claude Code plugin provides two agents for developing and executing content marketing strategies, conducting web research, and synthesizing information for marketing operations. It enables marketing teams to plan campaigns, analyze market trends, and organize content workflows without relying on external APIs or third-party services.

This Claude Code plugin manages conversation context by persisting and restoring conversation state across sessions, enabling long-running multi-turn interactions without losing track of prior exchanges. It provides commands and an agent interface for maintaining and retrieving conversation history and context information.

This plugin automates customer support workflows, sales pipeline management, and email campaigns through two agents. It integrates CRM functionality to streamline customer relationship management and sales processes.

This plugin provides database query optimization, cloud cost analysis, and scalability recommendations through four specialized agents that analyze database performance and infrastructure efficiency. It integrates directly into Claude Code without requiring external API keys or third-party services.

This plugin provides database architecture guidance, schema design patterns, and SQL optimization techniques for production systems through AI-assisted analysis and recommendations. It leverages two agents and one skill to help developers build performant and maintainable database solutions.

This database-migrations plugin automates schema migrations with built-in observability, supporting multiple database systems and cross-database migration strategies. It provides 2 commands and 2 agents to orchestrate and monitor migration workflows without requiring external API keys or tools.

This Claude Code plugin provides 2 commands and 2 agents with 4 supporting skills for constructing ETL pipelines, designing data warehouses, and building batch processing workflows. All functionality is self-contained with no external API keys or tool dependencies required.

This plugin provides schema validation, data quality monitoring, and streaming validation pipelines for backend APIs. It includes an agent-based validation system designed to handle input validation tasks for API endpoints and data quality assurance workflows.

This debugging toolkit plugin provides interactive debugging and developer experience optimization through 1 command and 2 agents that help developers identify and fix issues more efficiently. It integrates smart debugging workflows directly into Claude Code without requiring external APIs or tools.

This Claude Code plugin provides dependency auditing, version management, and security vulnerability scanning through an integrated command and agent interface. It identifies outdated packages and detects vulnerabilities without requiring external API keys or tools.

This plugin provides Claude Code with two agents that model deployment patterns, rollback automation, and infrastructure templates for safe application releases. It enables developers to plan and execute multi-stage deployments with automated recovery strategies.

This plugin provides pre-deployment checks and configuration validation to ensure deployment readiness before release. It includes one command and one agent that assess system state and configurations to catch issues early.

This Claude Code plugin bundles 11 essential developer skills covering Git workflows, SQL optimization, error handling, code review, end-to-end testing, authentication, debugging, and monorepo management through a single agent. It requires no external API keys or tools, making it a self-contained resource for core development practices.

This plugin provides operational tools for NVIDIA DGX Spark environments, including aarch64/CUDA-13 stack configuration, preflight checks for training gotchas, and management of unified memory and thermal systems. It bundles one command, one agent, and three skills to streamline local ML workload execution on Grace Blackwell hardware.

This Claude Code plugin provides distributed system tracing and debugging capabilities across microservices, implemented through 1 command and 2 agents. It enables developers to track requests and diagnose issues across service boundaries without requiring external API keys or tools.

This Claude Code plugin automates documentation generation by creating OpenAPI specifications, Mermaid diagrams, tutorials, and API reference documentation through a set of coordinated agents and skills. It requires no external dependencies or API keys, making it fully self-contained for developers who need to quickly document their code and APIs.

This plugin provides HADS (Human-AI Document Standard), a semantic tagging convention that makes documentation readable for both humans and AI models, along with the grounded-vault pattern for knowledge storage with per-claim provenance links and git fingerprints. It enables zero-token drift detection by tracking documentation changes at the claim level.

This plugin provides expert guidance and templates for building production-grade .NET applications using C#, ASP.NET Core, Entity Framework Core, and Dapper through a specialized .NET architect agent. It includes reusable patterns, best-practice templates, and comprehensive references covering API design, performance optimization, caching strategies, and testing approaches.

This plugin provides error analysis and trace debugging through 3 commands and 2 specialized agents that diagnose problems collaboratively. It requires no external APIs or tools, running entirely within Claude's environment.

This Claude Code plugin provides error tracing and root cause analysis for production systems through 3 commands and 2 agents. It enables smart debugging by analyzing error patterns and helping teams identify the source of issues in live environments.

This plugin provides a single skill for converting files between hundreds of formats (PDF to Word, HEIC to JPG, MP4 to MP3, CSV to JSON, EPUB to MOBI) using the free ChangeThisFile API. It requires no API key and offers MCP-aware implementation with a zero-dependency fallback script.

This plugin provides workflows for evaluating framework updates, planning migration strategies, and guiding architectural transformations through coordinated agents and skills. It helps teams assess migration complexity and execute framework changes systematically.

This plugin provides frontend UI development and mobile application implementation capabilities across platforms through 2 agents and 4 reusable skills. It supports cross-platform mobile development workflows without requiring external tools or API keys.

This plugin provides security best practices for frontend and mobile development, covering XSS prevention, CSRF protection, content security policies, and secure storage patterns. It includes one command and three agents to help developers implement these security measures in their applications.

This Claude Code plugin provides end-to-end feature orchestration across testing, security, performance, and deployment stages through a single command and four integrated agents. It consolidates quality assurance workflows without requiring external API keys or third-party services.

This Claude Code plugin provides two agents for functional programming education and development using Elixir, OTP patterns, and the Phoenix framework. It covers distributed systems concepts and practices without requiring external APIs or tools.

This plugin provides two agents and two skills for game development workflows: one agent handles Unity C# scripting and game mechanics, while the other manages Minecraft server plugin development using Bukkit/Spigot APIs. All functionality runs locally without external API dependencies or third-party services.

This plugin automates Git workflows and pull request processes through three commands and an agent, streamlining code review and team onboarding without external dependencies. It enhances PR management and collaborative development by integrating directly with Git operations.

Hermes Tweet provides Agent skills for public X/Twitter research via local discovery and API-driven read-only operations, with optional approval-gated access to private reads, writes, webhooks, and media operations. Setup requires the XQUIK_API_KEY environment variable and explicit enabling of state-changing operations through HERMES_TWEET_ENABLE_ACTIONS.

This plugin provides HR policy documentation, legal compliance templates covering GDPR, SOC2, and HIPAA standards, employment contracts, and regulatory documentation through integrated agents and skills. It enables organizations to generate, manage, and ensure compliance with employment law and data protection requirements without external dependencies.

This plugin provides production incident management capabilities with triage workflows and automated incident resolution through 6 agents and 3 skills. It enables teams to detect, prioritize, and resolve incidents directly from Claude Code using built-in commands and workflows.

A Claude Code plugin that enables JavaScript and TypeScript development with ES6+, Node.js, React, and modern web frameworks through 1 command, 2 agents, and 4 skills. It provides built-in support for contemporary JavaScript tooling without requiring external API keys or services.

This plugin provides comprehensive Julia development support including package management, scientific computing, and high-performance numerical code guidance within Claude Code environments. It delivers best practices for modern Julia 1.10+ projects through integrated agent-based assistance.

This plugin provides three AI agents for developing applications in JVM languages including Java, Scala, and C#, incorporating enterprise patterns and frameworks. The agents operate without external dependencies or API keys, making them self-contained development assistants.

This plugin provides Kubernetes manifest generation, networking configuration, security policies, observability setup, GitOps workflows, and auto-scaling capabilities through an agent and four skills. It requires no external API keys or tools, making it suitable for self-contained Kubernetes operations within Claude Code.

This Claude Code plugin provides production-ready LLM application development with LangGraph-based agents, advanced RAG systems, vector search optimization across multiple databases (Pinecone, Qdrant, Weaviate, Milvus, pgvector), and modern prompt engineering patterns. It includes three specialized agents (AI Engineer, Prompt Engineer, Vector Database Engineer) and eight skills covering architecture, evaluation, embedding strategies, and hybrid search implementation.

This plugin provides a complete LLM fine-tuning workflow supporting LoRA/QLoRA supervised fine-tuning, preference optimization methods (DPO/ORPO/KTO), reinforcement learning (GRPO/RLVR), vision model tuning, and quantized export — using Unsloth as the primary backend with TRL fallback. Fine-tuning operations are gated by evaluation results, meaning no evaluation harness prevents the fine-tune from proceeding.

This MLOps plugin provides tools for building ML model training pipelines with automated hyperparameter tuning and experiment tracking capabilities. It streamlines model deployment workflows through integrated agents and skills that orchestrate common machine learning operations without external dependencies.

This plugin orchestrates AI image generation workflows with parallel multi-direction output, prompt engineering, and access to a 1,300+ curated inspiration library through the MeiGen MCP server. It supports multiple providers (MeiGen Cloud, local ComfyUI, and OpenAI-compatible APIs) with free features including gallery search and prompt enhancement available immediately.

This Claude Code plugin coordinates cross-platform application development across web, iOS, Android, and desktop using 1 command and 6 specialized agents. It provides guidance and orchestration for building consistent implementations across multiple platforms without requiring external dependencies or API keys.

This plugin provides comprehensive observability infrastructure including metrics collection, logging, distributed tracing, SLO implementation, and monitoring dashboards through 4 agents and 4 reusable skills. It operates without external dependencies, making it suitable for self-hosted monitoring and observability workflows.

This plugin provides five session lifecycle and deployment agents that manage code review, deployment verification, and production monitoring using project-specific {{placeholders}}. It integrates session-start/end handlers with pre-ship code review, live deployment verification, and log health checks for a portable operating workflow.

This Claude Code plugin provides payment gateway integration with support for Stripe and PayPal, including checkout flow implementation, subscription billing, and PCI compliance features. It comprises one agent and four reusable skills for handling payment processing workflows.

This plugin provides performance analysis, test coverage review, and AI-powered code quality assessment through two integrated agents and commands. It analyzes code performance metrics and test coverage without requiring external APIs or tools.

A three-layer quality evaluation framework for Claude Code plugins using static analysis, LLM-based semantic evaluation, and Monte Carlo statistical testing to assess reliability and correctness. It provides CLI commands for scoring individual plugins, full certification with badges, and comparative head-to-head analysis.

This plugin creates editable, production-ready PowerPoint decks from a specification and source materials using a coordinate-explicit workflow with quality gates. It provides native shape, text, table, and image support while performing read-only analysis of reference decks and OOXML integrity validation.

This plugin enables modern Python development using Python 3.12+ with support for popular frameworks like Django and FastAPI, including async patterns and production best practices. It provides 3 agents and 16 skills to guide developers through architecture, testing, deployment, and code quality workflows.

This Claude Code plugin provides 2 agents and 2 skills for quantitative analysis, algorithmic trading strategies, and backtesting financial models. It enables portfolio risk management and financial modeling without requiring external API keys or integrations.

This plugin provides 3 agents and 4 skills for binary reverse engineering, malware analysis, firmware security audits, and software protection research in authorized security contexts. It supports CTF competitions and defensive security workflows without requiring external API keys or tools.

This plugin validates compliance with SOC2, HIPAA, and GDPR standards through automated scanning, secret detection, and regulatory checklists. It provides compliance validation and documentation generation without requiring external API keys or services.

This plugin performs static application security testing (SAST), scans dependencies for vulnerabilities, validates OWASP Top 10 compliance, and analyzes container security through 3 commands and 2 agents. It provides automated security hardening recommendations without requiring external API keys or tools.

This Claude Code plugin contains 3 agents that analyze SEO performance across content freshness, keyword cannibalization, and domain authority metrics. It requires no external API keys or integrations and runs entirely within Claude's environment.

This plugin provides three agents to handle SEO-focused content creation, planning, and auditing with E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) optimization. It operates entirely locally without external API dependencies or integrations.

This plugin analyzes and optimizes web pages for technical SEO, including meta tags, keywords, page structure, and featured snippet formatting. It contains four agents that assess and improve SEO elements to help pages rank better in search results.

This plugin provides production-grade Bash scripting capabilities with built-in defensive programming patterns, POSIX compliance validation, and comprehensive testing frameworks. It includes 2 agents and 3 skills for generating robust shell scripts and validating their quality across different environments.

Ship-mate orchestrates a full feature development pipeline from story to shipped code, routing work through specialized AI agents (orchestrator, architect, developer, PR reviewer, QA) and automated Playwright testing. It transforms a story file into reviewed, tested, and deployment-ready code by automating the entire development workflow.

A teaching skill that walks through setting up cryptographically signed audit trails on Claude Code tool calls using Ed25519 signatures, JCS canonicalization, and Cedar policies. Includes runnable examples, tamper detection walkthroughs, and CI/CD integration patterns as a precursor to the production-ready protect-mcp plugin.

This plugin provides three behavioral skills that detect and mitigate AI-specific failure patterns: hidden code debt from agent generation, context compaction amnesia, and imprecise Design Mode edits. The skills help maintain code quality and context integrity during AI-assisted development workflows.

This plugin schedules and publishes social media posts across 13 platforms including X, LinkedIn, Instagram, Facebook, TikTok, Discord, Telegram, YouTube, Reddit, WordPress, and Pinterest through a single SocialClaw API workspace key. It provides an agent interface for managing multi-platform social media distribution without requiring individual API keys for each service.

This Claude Code plugin provides expert-level startup business analysis with five specialized skills covering market sizing (TAM/SAM/SOM), financial modeling with scenarios, team planning, competitive analysis, and key metrics frameworks. It activates automatically based on business questions and includes three interactive commands plus one dedicated agent for comprehensive early-stage startup guidance.

This plugin teaches Claude agents to drive the Superself `self` CLI, which version-controls a project's state as an append-only event log in a separate git repository, enabling sessions to read context at start, attach work to units, report with evidence, and hand off to the next session. The skill wraps the free, local-first, open-source Superself CLI and issues shell commands for session management, work tracking, reporting, and decision logging.

This plugin provides multi-language support for systems programming with Rust, Go, C, and C++ through 4 specialized agents and 3 skills, enabling performance-critical and low-level development tasks. It includes 1 command and requires no external API keys or third-party services to operate.

This plugin implements test-driven development workflows with automated red-green-refactor cycles and integrated code review processes. It provides commands and agents to enforce TDD methodology throughout the development lifecycle.

This plugin automates team workflows, issue management, and standup processes within Claude, enabling developers to streamline collaboration without leaving their IDE. It includes 2 commands and 1 agent to optimize developer experience and coordinate team activities.

A comprehensive UI/UX design plugin providing guidance on mobile and web application design, including design systems, accessibility compliance (WCAG 2.2), and modern interaction patterns across iOS, Android, React Native, and web frameworks. It offers design reviews, component creation, accessibility audits, and design system setup through multiple agents and skills.

Automates unit and integration testing for Python and JavaScript codebases with integrated debugging capabilities. The plugin provides 1 command and 2 agents to help write, run, and troubleshoot tests within your development workflow.

This plugin provides two agents for web scripting using PHP and Ruby to develop web applications, CMS systems, and backend services. It enables developers to build and manage server-side logic without requiring external API keys or tools.

This plugin provisions direct cloud access for the Claude Office add-in by configuring Vertex AI, Bedrock, Azure AI Foundry, or an LLM gateway backend, then generates a customized add-in manifest and handles Azure admin consent workflow. It also manages per-user configuration via Microsoft Graph extension attributes and supports dynamic MCP server bootstrapping and user data export.

This plugin processes earnings calls and SEC filings to extract key financial data and automatically update financial models with extracted insights. It organizes findings into structured notes and drafts for further analysis.

The equity-research plugin provides tools for financial analysis workflows including earnings analysis, initiating coverage reports, and structured research output. It includes 9 skills and commands to streamline equity research processes.

This Claude Code plugin provides financial modeling and analysis capabilities including DCF valuation, comparable company analysis, LBO models, three-statement financial modeling, competitive analysis, and pitch deck quality checks. The plugin contains 7 commands and 13 skills that work together to support end-to-end financial analysis workflows without requiring external APIs or third-party services.

This plugin provides six finance operations skills for fund administration, including general ledger reconciliation, break tracing, accruals management, roll-forwards, variance commentary, and net asset value tie-out. All capabilities are self-contained within the plugin with no external dependencies or API requirements.

gl-reconciler finds breaks in general ledger reconciliation, traces their root cause, and routes exceptions for sign-off through a workflow. It bundles one agent and four skills to automate the investigation and approval of accounting discrepancies.

This plugin provides investment banking productivity tools for creating deal materials (CIMs, teasers, buyer lists), presentations (strip profiles, pitch decks), and transaction analysis (merger models, deal tracking). It operates entirely through Claude commands and skills with no external dependencies or third-party API integrations.

kyc-screener is a Claude Code plugin that parses onboarding documents and runs a rules engine to automatically flag compliance gaps during KYC (Know Your Customer) workflows. It uses one built-in agent and three supporting skills to validate documentation against regulatory requirements without requiring external services.

market-researcher is a Claude Code plugin that transforms a sector or theme into a comprehensive industry overview, competitive landscape analysis, peer comparisons, and actionable ideas shortlist. It uses 1 agent and 5 supporting skills to synthesize market research without requiring external APIs or tools.

This Claude Code plugin prepares comprehensive briefing packs for client meetings by gathering and organizing relevant information through an integrated agent. It uses four skills to synthesize meeting-specific context and recommendations before each engagement.

This Claude Code plugin enables financial modeling directly within Excel, supporting DCF (Discounted Cash Flow), LBO (Leveraged Buyout), 3-statement, and comps analysis workflows. It comprises one agent coordinating six specialized skills to build and manipulate these common financial models live in your spreadsheet.

This Claude Code plugin automates month-end accounting workflows including accruals, roll-forwards, and variance commentary generation. It provides one agent with five supporting skills to streamline financial close processes.

This plugin provides two skills for operational workflows: parsing KYC (Know Your Customer) documents and evaluating business rules within a grid-based framework. It enables document processing and rule validation without requiring external APIs or services.

This plugin guides users through financial analysis workflows—comparing competitor valuations, analyzing precedent transactions, and structuring leveraged buyout scenarios—before converting all findings into a cohesive branded pitch deck. It bundles eleven skills within a single agent to automate end-to-end deal preparation for investment professionals.

This plugin provides command-driven tools for private equity deal sourcing, including company discovery, CRM integration workflows, and automated founder outreach capabilities. It streamlines the investment workflow by organizing deal research and prospect management within Claude Code.

This plugin provides an agent with three skills to audit pre-generated LP (limited partnership) statements before they are distributed, checking for accuracy and compliance issues. The auditing capabilities operate entirely within Claude Code with no external dependencies or third-party integrations.

This plugin ingests GP (General Partner) packages, executes valuation templates, and stages data for LP (Limited Partner) reporting workflows. It bundles one agent and four supporting skills to automate the valuation review pipeline.

This plugin scaffolds new Claude Agent SDK applications in Python and TypeScript with the `/new-sdk-app` command, automatically installing the latest SDK version and running language-specific verifier agents to ensure best practices compliance. The included verifier agents thoroughly check SDK setup, configuration, environment security, and code patterns for both Python and TypeScript projects.

This plugin connects Claude Code to Asana's V2 MCP server, enabling task creation, project search, assignment updates, and progress tracking directly from the terminal. Users must complete a one-time OAuth app setup (via the /asana-setup command) to authenticate with their Asana workspace.

This plugin integrates clangd, the LLVM language server, into Claude Code to provide code intelligence, diagnostics, and formatting for C/C++ projects. It enables IntelliSense-like features and error checking across common C/C++ file extensions (.c, .h, .cpp, .cc, .cxx, .hpp, .hxx, .C, .H).

This skill analyzes codebases to recommend tailored Claude Code automations including hooks, skills, MCP servers, and subagents that match the project's specific needs. It operates in read-only mode, scanning files to identify the top 1-2 automation recommendations in each category.

The claude-md-management plugin provides tools to audit CLAUDE.md files against codebase state and capture session learnings into project memory. It includes a skill for periodic maintenance audits and a command for end-of-session documentation updates.

The Code Review Plugin automates pull request review by launching four parallel agents that independently audit changes from different perspectives—CLAUDE.md compliance, bug detection, and git history analysis—and score each issue for confidence to filter false positives. Only high-confidence issues (≥80 score) are posted as review comments, reducing noise while ensuring actionable feedback.

The code-simplifier agent takes existing code and refactors it for improved clarity, consistency, and maintainability without altering its functionality. It works entirely using Claude's built-in capabilities with no external dependencies or third-party service integrations.

This Claude Code plugin provides three git workflow commands: `/commit` analyzes staged and unstaged changes to generate context-aware commit messages, `/commit-push-pr` orchestrates commits, pushes, and PR creation in one step, and `/clean_gone` removes local branches deleted from the remote repository. The plugin requires Git to be installed, and `/commit-push-pr` additionally requires GitHub CLI (`gh`) to be authenticated.

This Claude Code plugin integrates the csharp-ls language server to provide code intelligence, diagnostics, and editor features for C# files (.cs). It requires .NET SDK 6.0 or later and can be installed globally via dotnet tool or Homebrew.

This Claude Code plugin provides seamless onboarding for the Code-with-Claude Makers M5Stack Cardputer-Adv, automatically cloning the build repository, flashing UIFlow firmware, and installing the Claude Buddy app bundle with a single `/maker-setup` command. After setup, the plugin enables iterative app development via MicroPython, allowing you to push and test new applications directly to the device without re-flashing.

The Feature Development Plugin provides a structured 7-phase workflow for building new features, including codebase exploration, architecture design, clarifying questions, and quality review using specialized agents. It guides developers through understanding requirements, designing approaches, implementing code, and reviewing for quality before integrating features into existing codebases.

This plugin enables Claude to generate production-grade frontend interfaces with distinctive aesthetics, bold typography, custom color palettes, and high-impact animations instead of generic AI-generated designs. Claude automatically applies this skill when working on UI/UX tasks, choosing a clear aesthetic direction and implementing meticulous production code.

gopls-lsp is a Go language server that integrates with Claude Code to provide code intelligence, refactoring, and analysis for Go files. It requires installation of gopls via the Go toolchain and relies on standard Go development tools.

This plugin integrates Eclipse JDT.LS (Java Language Server) with Claude Code to provide code intelligence, diagnostics, and refactoring capabilities for Java files. It enables IDE-like features such as autocompletion, error detection, and code navigation for `.java` files.

This Kotlin language server plugin for Claude Code provides code intelligence, refactoring, and analysis for .kt and .kts files by integrating the kotlin-lsp tool. It requires installation of the Kotlin LSP CLI via Homebrew.

A Lua language server that provides code intelligence, diagnostics, and editor support for `.lua` files within Claude Code. It integrates as a language server protocol (LSP) implementation to enable real-time linting, autocompletion, and documentation for Lua development.

Solves competition mathematics problems (IMO, Putnam, USAMO) using context-isolated adversarial verification that catches self-verification bias by testing proofs with specific failure patterns. Calibrates abstention over false claims and produces clean LaTeX/PDF output after verification passes.

Two Claude Code skills tackle hard mathematics problems: `/math-proof:solo` works through a problem in one session with intermediate results saved to a notes file, while `/math-proof:siege` runs rounds of judge and worker sub-agents over hours to collaboratively prove or refute the goal, ending with a plainly-stated `proof.md` documenting what was and was not proved.

This plugin provides three skills to guide developers through designing and building MCP servers for Claude, covering deployment models (remote HTTP, MCPB, local stdio), tool design patterns, and interactive UI creation. The entry skill interrogates the use case and routes to specialized skills for building remote servers, in-chat applications with widgets, or packaged local binaries.

This plugin automates the setup of an Anthropic MCP tunnel, guiding you through certificate generation, Docker Compose configuration, and cloudflared proxy deployment to securely expose a private MCP server to Claude over an outbound-only connection. It includes a verifiable FastMCP sample server and provides troubleshooting guidance for TLS, configuration, and operational tasks like token rotation.

This Claude Code plugin integrates Intelephense, a PHP language server, to provide code intelligence, diagnostics, and IDE-like features for PHP files. It enables real-time linting, type checking, and other language analysis capabilities when working with `.php` files in Claude Code.

Playground plugin creates self-contained interactive HTML files with visual controls, live preview, and copy-able prompt output for topics with large or complex input spaces. It includes templates for design decisions, data exploration, concept mapping, and document review workflows.

This plugin provides seven specialized skills and a structured workflow command for building high-quality Claude Code plugins, covering hooks, MCP integration, plugin structure, settings, commands, agents, and skills. It includes comprehensive documentation, working examples, and utility scripts to guide developers through an 8-phase plugin creation process from discovery to marketplace publishing.

Generates and publishes a living project status page as a tabbed HTML artifact on claude.ai, tracking workstreams, success criteria, and progress over time with support for PR-driven software projects. The page persists at a shareable URL and can be refreshed in later sessions to reflect live state from repositories, trackers, or design docs.

This plugin bundles 6 specialized agents for pull request review, each focusing on a distinct aspect: code comments, test coverage, error handling, type design, code quality, and simplification. They analyze PRs individually or in combination to provide comprehensive code quality feedback with actionable, severity-ranked findings.

A Python language server (Pyright) integration for Claude Code that enables static type checking and code intelligence for .py and .pyi files. It requires Pyright to be installed globally via npm, pip, or pipx, and leverages Microsoft's Pyright type checker for real-time Python analysis.

Ruby LSP integrates the Ruby language server into Claude Code, providing syntax analysis, code intelligence, and real-time diagnostics for Ruby files (.rb, .rake, .gemspec, .ru, .erb). It requires Ruby 3.0+ and can be installed as a gem for IDE-level code support.

rust-analyzer-lsp is a language server that enables code intelligence and analysis for Rust files in Claude Code. It provides real-time linting, diagnostics, and IDE-like features for .rs files by integrating the rust-analyzer tool.

This Claude Code plugin enables users to create new skills from scratch, improve and optimize existing skills, and measure performance through evaluations and benchmarking. It provides tools for skill development, testing, and variance analysis to ensure reliable performance.

Provides code intelligence for Swift projects in Claude Code by integrating SourceKit-LSP, the official Swift language server. Enables syntax highlighting, code completion, and diagnostics for .swift files without additional configuration beyond the Swift toolchain.

A TypeScript/JavaScript language server for Claude Code that provides code intelligence features including go-to-definition, find references, and error checking across .ts, .tsx, .js, .jsx, .mts, .cts, .mjs, and .cjs files. The plugin requires the typescript-language-server and typescript packages to be installed globally via npm or yarn.

This plugin performs AI-powered code reviews on Git diffs, supporting workspace changes, branch ranges, and individual commits with concurrent per-file analysis and codebase search capabilities. It provides context-aware review feedback without requiring external API keys or tools.

This is the Claude Code Plugins Directory documentation itself, explaining the structure, installation, and contribution guidelines for the plugin marketplace. It is not a functional plugin but rather the informational hub for discovering, installing, and developing Claude Code plugins.

Codex plugin for Claude Code integrates OpenAI's Codex directly into Claude Code workflows, offering `/codex:review` and `/codex:adversarial-review` commands for code reviews plus job management commands to delegate tasks and hand off sessions to Codex. It requires a ChatGPT subscription or OpenAI API key and Node.js 18.18+, routing usage through Codex's own limits.

A curated library of 500+ reverse-engineered prompts and 20+ industrial templates for GPT-Image 2 image generation, organized by use case and style with a searchable web gallery. The plugin includes two agents that help users explore, refine, and apply these prompts to create images using the GPT-Image 2.5 Sunburst and Flare models.

agent-flow is a Claude Code plugin that renders a live visualization of the session's agent tree in a side pane, showing the main loop's context window, spawned subagents, and the context/work/answers flowing between them. It observes session events to track agent spawns, model requests, tool calls, and loop completions without modifying them.

diff-invaders is a Space Invaders-style game plugin for Claude Code where alien formations are generated from code diffs—each line of added code becomes a wave row, and every two characters spawn one alien. Players defend against these diff-based waves with a ship, single shots, and bombs across three lives in an interactive terminal.

jev-skill-suggestion reduces context bloat by using TypeSafe's Jev System One model to intelligently select at most one relevant skill per prompt, then injects only that skill's documentation rather than listing all skills to the model. Falls back to Claude's built-in classifier when no API key is configured.

A Claude Code plugin that displays a live prompt-cache meter showing token read/write/uncached counts, countdown to cache expiry, and actionable recommendations (keep going, /compact, or /clear). It infers your account's cache TTL (5 minutes or 1 hour) from traffic patterns and Claude Code settings, and surfaces cache misses with their cause.

secret-redactor intercepts tool results and removes credential-shaped strings (API keys, tokens, JWTs, private keys, connection strings) before the model sees them, while preventing redacted placeholders from being echoed into Bash commands. It uses a tool.call hook in the "after" placement to deep-replace secrets and notify the model of redactions via context.

tool-defense is a tower defense game embedded in Claude Code where enemies represent real tool calls Claude makes—Bash spawns fast runners, Edit/Write create slow tanks, web or MCP calls spawn flying enemies, and Agents spawn bosses. The game runs entirely client-side in the interactive terminal with zero token cost, requiring you to build and upgrade towers to defend against Claude's tool usage waves.

frontend-slides is a zero-dependency HTML presentation generator that creates professional slideshows with bold template-library styles and built-in PowerPoint conversion. It emphasizes clean, anti-AI-slop design principles and requires no external services or API keys.

This plugin enables teams to create, customize, and manage Claude Code plugins directly within their organization's ecosystem, including configuration of plugin behavior and template adaptation. It provides two skills for managing plugin lifecycle and workflows without requiring external API keys or tools.

agentmemory provides persistent memory for AI coding agents (Claude Code, Cursor, Gemini CLI, and others), allowing agents to recall context and decisions across sessions without re-explanation. Built on the iii engine, it offers 95%+ retrieval accuracy with 92% fewer tokens and works with any MCP client, supporting 54 MCP tools with zero external database dependencies.

Claude HUD displays real-time context usage, tool activity, agent status, and todo progress in a native statusline below your input, with no separate window or external dependencies. It pulls native token data from Claude Code and re-renders after each interaction, offering customizable presets and multi-language support.

This plugin scans frontend projects for 20+ WCAG 2.2 accessibility violations across React, Next.js, Vue, Angular, Svelte, and HTML, with a stdlib-only contrast checker and framework-specific fix patterns. It runs via command-line with CI-friendly exit codes and requires no external dependencies.

This AEO (Answer Engine Optimization) plugin optimizes content to be cited as authoritative sources by AI language models like ChatGPT, Perplexity, Claude, Gemini, and Mistral, distinct from traditional SEO. It includes three Python tools for E-E-A-T scoring, content rewriting in multiple modes, and local citation tracking, with industry-specific thresholds for SaaS, healthcare, finance, legal, ecommerce, B2B, media, and education.

agent-harness provides a bounded agentic loop that compiles goals into verified task plans, executes with retry budgets and controller-run verification, and escalates to humans when budgets exhaust rather than faking success. It ships a manifest builder, goal compiler, loop controller state machine, 18 per-domain harness manifests, and a /cs:harness command for end-to-end goal execution.

AgentHub spawns N parallel agents in isolated git worktrees to compete on optimization, refactoring, content generation, or research tasks, then evaluates results by metric or LLM judge and merges the winner. It includes 7 slash commands, agent templates (optimizer, refactorer, test-writer, bug-fixer), and git DAG orchestration with a message board for coordination.

The agile-product-owner Claude Code plugin generates INVEST-compliant user stories with multiple acceptance criteria patterns (Given/When/Then, rule-based, checklist) and breaks epics into deliverables using five split techniques. It provides sprint planning with velocity-based capacity math and weighted backlog prioritization powered by a user_story_generator Python tool.

This Claude Code plugin channels Marc Andreessen's market-first decision framework to pressure-test ventures, ideas, features, and career bets through documented principles (market dominates; product/market fit is the only milestone that matters; bias to build), issuing hard verdicts backed by deterministic scoring tools. It also runs Andreessen's daily 3x5 index card and Anti-Todo productivity routine with an anti-sycophancy operating prompt that leads with counterarguments, avoids hedging, and provides explicit confidence levels on all claims.

This plugin guides developers through Apple's Human Interface Guidelines with specialized knowledge of 2026 Liquid Glass aesthetics, offering design validation and auditing for iOS, macOS, and visionOS applications. It includes a hig_checker Python tool that validates tap targets, contrast ratios, and accessibility compliance across Apple platforms.

Guides founders through a structured 12-phase business planning interview (foundation, strategy, market, financial, sales, marketing, product, operations, tech, people, legal, governance), generating concepts as version-controllable markdown files organized in an Open Knowledge Format bundle. Includes built-in validation and scaffolding tools to ensure conformance and index consistency.

An autonomous experiment loop agent that optimizes any file by a measurable metric, with configurable loop intervals ranging from 10 minutes to monthly. It provides 8 evaluators and 6 skills to measure and improve file performance iteratively.

This plugin implements a Self-Mirror consciousness loop technique that injects inner dialogue into AI responses, creating a three-step process (Self → Mirror → Conscious Response) to generate more human-like and authentic output. It requires no external dependencies or API keys, relying entirely on prompt engineering with a single Agent Skill.

Converts books and documentation (PDF, EPUB, DOCX, HTML, Markdown, RST, AsciiDoc, RTF, MOBI/AZW) into structured agent skills with a resident core framework index, on-demand chapter files, glossary, patterns, and decision cheatsheet. Ships four stdlib-only tools for extraction, validation, token budgeting, and plugin emission, plus an agent and commands to compile sources into reusable knowledge bases.

This plugin provides five business and growth skills—customer success manager, sales engineer, revenue operations specialist, and contract writer—plus a router to distribute tasks among them. All capabilities are self-contained with no external API dependencies, making them suitable for Claude Code, Codex, Gemini CLI, Cursor, and OpenClaw environments.

This Claude Code plugin provides business investment analysis and capital allocation advice for evaluating expenditures on equipment, real estate, new ventures, hiring, and technology. It operates as a standalone skill with no external dependencies or API integrations required.

This plugin provides 6 specialized business operations skills—process mapping with BPMN bottleneck analysis, vendor management scorecards, capacity planning via Erlang-C queueing theory, change communication frameworks, operational runbook authoring, and procurement spend optimization—routed through an orchestrator agent that uses context: fork to direct inquiries to the appropriate sub-skill. All 18 underlying Python tools rely on stdlib only, with 24+ reference documents (7+ authoritative sources per skill) and reusable asset templates for each domain.

capture transforms unstructured brain dumps of mixed thoughts, tasks, and ideas into a clean four-section actionable system (Projects/Ideas, Tasks, Connections, How I Can Help) with zero information loss. It asks at most one clarifying question when genuinely ambiguous, uses real workspace detection (Glob/Grep verified), and defaults to compressed output for small dumps.

Ultra-compressed communication mode that cuts token usage ~75% by eliminating filler, articles, and pleasantries while preserving technical accuracy. Includes Python tools for text compression, token-savings estimation, and caveman-style linting, plus a persistent agent and slash command for activation.

This plugin provides end-to-end chaos engineering discipline through three Python tools for designing experiments with hypothesis-driven testing, calculating blast radius against error budgets, and generating blameless postmortems. It includes reference documentation on chaos principles and attack taxonomy, experiment templates, and an interactive slash command to guide users through the full chaos engineering workflow.

A standalone skill for Chief AI Officer-level strategic decision-making that provides build-vs-buy calculators, EU AI Act and US state regulation classifiers with 7+ citation-backed analyses, and 2026-projected AI cost economics across self-hosted and API models. Designed as a strategic complement to engineering-focused AI skills, with deep references anchoring each major advisory domain to 5+ authoritative sources.

This skill guides Chief Customer Officer strategic decisions on retention decomposition (GRR vs NRR analysis), customer segmentation design (4-tier framework with ICP scoring), and CS coverage planning (pooled vs named CSM models with 12-month hiring projections). It covers only high-level strategic advisory for B2B SaaS retention and team organization, excluding tactical customer success execution.

This plugin provides Chief Data Officer advisory across four strategic decisions: AI training data audit (with GDPR and EU AI Act compliance mapping), data product strategy selection (warehouse vs. lakehouse vs. mesh with 12-month roadmap), data asset valuation (strategic scoring and M&A multipliers), and data team organizational evolution. It operates entirely on standard library with no external dependencies or API keys required.

claude-coach is a personal Claude power-user coach that delivers a personalized cheat-code glossary on first use, then scans subsequent prompts to surface at most one high-impact tip when genuinely helpful. It includes a coaching decision tree, prompt rater, and three stdlib Python tools bundled as a Claude Code skill with a dedicated slash command and agent persona.

A virtual C-suite plugin for Claude Code providing 13 cs-* agents (CFO, CMO, CRO, CPO, COO, CHRO, CISO, Chief of Staff, General Counsel, Chief Data Officer, Chief AI Officer, Chief Customer Officer, VP of Engineering) and 21 slash commands for founder-mode executive decision-making, multi-role boardroom deliberation, and strategic sprint workflows. The plugin wraps 22 underlying skills with cognitive gearing and artifact-driven pipelines, enabling forcing-question office hours, cross-model consensus reviews, and two-layer memory decision logging.

This plugin provides 34 C-level advisory skills covering CEO and CTO strategic decision-making, including tools for technical debt analysis, financial scenario modeling, team scaling, and board governance frameworks. The skills are designed for startups and enterprises making executive decisions on strategy, technology, and organizational leadership.

Analyzes frontend, backend, and fullstack codebases to automatically generate complete Product Requirements Documents, extracting routes, components, models, APIs, and data relationships. Uses Python scripts to scan code and scaffold structured PRD output with per-page documentation, API inventory, and enum dictionaries.

This Claude Code plugin generates CodeTour .tour files with persona-targeted step-by-step walkthroughs that link to specific files and line numbers, supporting 10 developer personas and all CodeTour step types. It includes built-in SMIG description formula support for structured documentation.

Collab-proof is a retrospective analysis tool that uses Claude to assess developer-versus-assistant contributions across a session using a 4-frame evaluation framework. It requires no external dependencies or API keys, analyzing collaboration dynamics entirely through Claude's built-in capabilities.

This plugin provides 7 commercial skills plus an orchestrator for pricing strategy, deal desk review, partnership tiering, channel economics, commercial policy, RFP response, and quarterly forecasting—all using stdlib-only Python tools and structured methodologies like Van Westendorp WTP analysis and Shipley method RFP evaluation. The orchestrator routes inquiries via context-forking and applies grill discipline to validate assumptions against SaaS pricing canon.

Compliance OS is a meta-orchestrator that helps teams manage multiple compliance frameworks in parallel by selecting applicable frameworks based on company profile, computing cross-framework overlap, simulating internal audits, and generating unified evidence checklists. It coordinates with 14 existing ra-qm-team skills and 2 new compliance plugins to handle ISO 27001, 13485, 42001, 14971, EU AI Act, MDR 745, GDPR, SOC 2, and FDA QSR across security, medical devices, AI governance, and privacy domains.

This plugin provides three deterministic tools for EU AI Act (Regulation (EU) 2024/1689) compliance: an AI system risk classifier (Article 5–50 per risk tier), a conformity assessment planner (Module A/H routing per Article 43 + Annex IV checklists), and an obligation tracker (provider/deployer/importer/distributor matrix with GPAI Articles 51–55). It is purpose-built for compliance officers executing Article-level conformity work and includes four references with Article-by-Article walkthroughs, high-risk categories, GPAI thresholds, and mappings to ISO 42001 + NIST AI RMF + GDPR.

This standalone ISO/IEC 42001:2023 AI Management System specialist skill helps compliance teams run internal AIMS audits through three tools: a gap analyzer for Clauses 4–10 coverage scoring, an AI risk register builder mapping controls to Annex A, and an audit scheduler for 12-month planning with auditor independence checks. It requires no external APIs or dependencies and provides 4 reference guides covering clauses, controls, maturity models, and cross-framework alignment with the EU AI Act, NIST AI RMF, and ISO 23894.

This Claude Code plugin provides three data quality assessment tools that audit datasets for completeness, consistency, accuracy, and validity using pure Python: a data profiler with DQS scoring, a missing value analyzer with MCAR/MAR/MNAR classification, and a multi-method outlier detector. All functionality relies on Python standard library only, requiring no external API keys or services.

A dated companion to the 2016 *Deep Learning* textbook that synth­esizes each chapter with a 2016–2026 delta layer tracking what held, what shifted, and what became historical, plus four stdlib tools for reading-path planning, training diagnostics, capacity planning, and parameter arithmetic. All four tools are executable, deterministic, and make no external calls—designed for studying or teaching the book and validating whether its recommendations remain current.

deepread analyzes supplied documents using five reading modes (quick, deep, map, feynman, book) and decomposes content into claims, reasons, and evidence with confidence labels. It generates evidence ledgers and knowledge maps to support comprehensive, evidence-first document comprehension without requiring external data sources.

Deep-research runs a disciplined 9-phase investigation pipeline that reframes questions into falsifiable hypotheses, fans out parallel searches across multiple source types, and triangulates every claim against at least 3 independent, differently-typed sources with verbatim quotes saved per-source. Output is an auditable, reusable folder with mandatory adversarial critique and a delta-update refresh protocol, designed for high-stakes decisions where accuracy matters more than speed.

The deep-work plugin classifies tasks as deep or shallow, audits the shallow-work share against a configurable budget, and builds an energy-optimized time-blocked day with deep blocks of at least 90 minutes in early hours (capped at 4 hours), shallow work batched into at most two windows, and attention-residue buffers. A local JSON ledger tracks weekly deep hours versus target and consecutive-day streaks, while deterministic Python scripts enforce refusals (exit code 2) when the plan exceeds budget or capacity.

This Claude Code plugin orchestrates Playwright, FFmpeg, and Edge TTS to create polished demo videos from screenshots and scene descriptions, complete with story structure and narration guidance. It generates product walkthroughs, feature showcases, and marketing teasers by composing automation, video encoding, and text-to-speech capabilities.

This Claude Code skill provides guidance on Dockerfile optimization, docker-compose orchestration, multi-stage builds, and container security hardening. It helps developers improve build performance, layer caching, and implement production-ready container patterns.

Produces a hypothesis-tested dossier on companies, people, nonprofits, or government organizations by forcing users to state their upfront thesis and allocating 30%+ of research to disconfirming evidence. Outputs an editable Word document with verdict, identity facts, 12-month timeline, network and reputation signals, red flags, conversation hooks, and a source-provenance audit log sourced from WebSearch, WebFetch, and free APIs (SEC EDGAR, GitHub, ProPublica).

A workflow-pair plugin implementing personalized recurring email triage: inbox-setup runs once to interview the user and build a 7-file knowledge base (taxonomy, patterns, evaluation framework, rate-card, blocklist, tracker, triage-log), while inbox-triage runs 1–3 times daily or on demand to classify recent emails, draft replies without sending, and update the knowledge base. Both skills share a strict file contract verified by cross-skill consistency audit.

A comprehensive suite of 39 advanced engineering skills covering system design, deployment pipelines, observability, and operational excellence. Includes agent designers, RAG architects, database schema designers, CI/CD builders, SLO architects, chaos engineering, and pre-production audit gates.

A comprehensive collection of 34 production-ready engineering skills covering architecture, frontend, backend, fullstack, QA, DevOps, security, AI/ML, and data engineering, each augmented with forcing-question libraries and deterministic decision engines. The plugin provides orchestrator agents (/cs:fullstack-review, /cs:frontend-review, /cs:backend-review, /cs:engineer-grill) for composition into specialists and works across Claude Code, Codex, Gemini CLI, Cursor, and OpenClaw.

An adversarial thinking partner that stress-tests business plans, prepares founders and executives for board meetings, helps navigate hard decisions, and conducts honest post-mortems. The plugin uses a single agent with six supporting skills to challenge assumptions and identify potential risks in strategic thinking.

This skill transforms rambling descriptions of desired outcomes into polished, autonomous /goal prompts ready to paste into a fresh Claude session. It extracts key requirements (deliverable, quantity, stakes, tools, destination), verifies resources against the live environment, and generates a structured 150–350 word prompt with seven anatomical parts before self-checking against six binary criteria.

This skill classifies, ships, ramps, and retires feature flags as a controlled lifecycle with measurable debt, providing Python tools for flag debt scanning, phased rollout planning, and kill-switch auditing. It includes 4 reference documents on flag taxonomy, provider comparisons, rollout strategies, and lifecycle management, plus a /flag-cleanup slash command for quarterly cleanup workflows.

The plugin provides financial analyst and SaaS metrics coach skills with capabilities for ratio analysis, DCF valuation, budgeting, forecasting, and SaaS KPI tracking (ARR, MRR, churn, CAC, LTV, NRR, Quick Ratio). It includes 7 Python automation tools and a nested business-investment-advisor skill for use in Claude Code, Codex, Gemini CLI, Cursor, and OpenClaw.

This plugin provides automated contract risk scanning (flagging founder-killer patterns like auto-renew traps and uncapped indemnity) and term sheet analysis (scoring founder-friendliness across 12 dimensions including liquidation preference and anti-dilution). It includes reference materials on startup contracts, IP/regulatory landscapes (HIPAA, GDPR, FDA, EU AI Act, SOC 2), and term sheet decoding, but explicitly disclaims legal advice.

This Claude Code plugin provides authentication and command automation for Google Workspace services (Gmail, Drive, Sheets, Calendar, Docs, Chat, Tasks) through the gws CLI, with 5 Python tools, 3 reference guides, 43 recipe command templates, and 10 persona bundles to generate and verify administrative commands.

A strategic NIH grant discovery and positioning plugin for clinical researchers that conducts a 5-facet Consensus analysis, maps research to the right NIH institutes via RePORTER, and produces an editable Word document with mechanism recommendations, timelines, and program officer guidance. NIH-only scope; non-NIH funders are flagged at intake and out of scope.

A relentless plan-and-design interrogator that walks decision trees one branch at a time, asking forcing questions sequentially with recommended answers, and explores the codebase to resolve answers where possible. Enhanced from Matt Pocock's MIT-licensed grill-me with stdlib Python tools for decision-tree extraction, question generation, and session-state tracking, plus reference docs on forcing-question patterns and a dedicated persona agent.

Guides decision-making sessions for architectural plans by cross-referencing them against a project's documented terminology (CONTEXT.md) and architectural decisions (docs/adr/), then updates those files inline as decisions crystallise. Includes three Python linting and consistency-checking tools plus Matt Pocock's original interview discipline preserved verbatim under MIT.

A conversation-handoff document generator that compacts the current session into a markdown brief for a fresh agent to continue, referencing existing artifacts (PRDs, plans, ADRs, issues, commits) by path/URL without duplication. Includes three stdlib Python tools (template generator, artifact deduplicator, skill recommender), a cs-handoff-author persona agent, and a /cs:handoff slash command.

This plugin provides an agent skill for Kubernetes Helm chart development, helping scaffold charts, design values files, implement template patterns, manage dependencies, harden security configurations, and run chart tests. It integrates with Claude Code, Codex, Gemini CLI, Cursor, and OpenClaw to streamline Helm workflow automation.

Builds a single-file HTML review page for Markdown or HTML artifacts that collects structured batch feedback as a sidecar instead of chat prose, and enforces a gate that refuses to close until all blockers are resolved, the reviewer is named, and integrity checks pass. The review page makes zero network requests and sanitizes HTML to prevent execution; feedback is hand-writable Markdown so the loop works over SSH and in CI where no browser exists.

Enforces Andrej Karpathy's 4 coding principles (surface assumptions, simplicity, surgical changes, verifiable goals) via 4 stdlib-only Python tools (assumption_linter, complexity_checker, diff_surgeon, goal_verifier), a review agent, and a /karpathy-check slash command. Integrates as a pre-commit gate to catch over-engineering, hidden assumptions, diff noise, and vague plans before they ship.

This plugin provides end-to-end discipline for building Kubernetes Operators, including three stdlib Python linting tools (CRD validator, reconcile-loop linter, capability auditor), reference documentation on operator patterns and framework comparisons, and production-ready CRD and Go controller templates. The `/operator-audit` slash command runs all tools together to catch common reconcile-loop bugs like missing finalizers, blocking calls, and status drift before they reach a cluster.

This Claude Code plugin generates premium single-file HTML landing pages with GSAP 3D animations, scroll-triggered reveals, and mouse-parallax effects, using a structured intake flow to lock down product positioning, audience, branding, and tone before output. It outputs a fully self-contained .html file with inline CSS/JS and only CDN externals (Google Fonts and GSAP), distinct from the conversion-optimized Next.js landing-page-generator.

This plugin enforces LinkedIn's User Agreement §8.2 at the code level, blocking automation, scraping, and engagement pods, then routes to five specialized skills for profile optimization, strategy validation, content linting, ethical engagement, and honest analytics. Every tool is pure Python (stdlib only), holds no credentials, makes no API calls, and returns text the account holder manually posts—treating LinkedIn presence as a months-long authentic asset rather than a posting schedule.

Converts a research question into a strategically planned mini literature review delivered as a Word document, using free keyless PubMed and OpenAlex APIs with optional Consensus MCP enhancement. Output is an 8-section orientation guide designed as a launching pad for researchers entering an unfamiliar field, with configurable depth (5/10/20 queries) and built-in integrity checks for sequential execution and source tracking.

This Claude Code plugin provides tools to reduce LLM API costs by implementing token optimization, model routing based on cost/quality tradeoffs, and prompt caching strategies. It enables cost observability and spending controls for AI-powered features without requiring external API keys or services.

This plugin turns Claude Code + Obsidian into a persistent wiki where knowledge compounds through controlled ingest workflows rather than being re-derived by RAG on every query. It ships a master SKILL, 3 sub-agents, 5 slash commands, and 8 stdlib Python tools to read sources once, build interlinked entity/concept/source pages, flag contradictions, and maintain a living index.

Converts long markdown files into single-file, self-contained interactive HTML with five specialized converters: long-form documents with sticky TOC and search, code reviews with 2-column diffs and annotations, and slide decks with presenter mode and keyboard navigation. A one-time design-system onboarding captures your brand palette and typography, ensuring consistent rendering across all conversions with WCAG-AA contrast enforcement.

A 48-skill marketing suite providing frameworks and automation tools for content creation, SEO, CRO, paid channels, growth, intelligence, and sales enablement across tech company go-to-market functions. The collection includes brand voice analysis, SEO optimization, demand generation playbooks, and campaign analytics with no external API dependencies.

A Claude Code plugin that enforces meeting discipline through three stages: a cost-gate that prices meetings in real dollars and refuses ones without a decision, agenda, and owner; a timeboxed agenda builder that orders decisions first and rejects topics without desired outcomes; and an action-item extractor that flags orphaned tasks and missing due dates. All three stages run via standard library scripts with no external service dependencies.

This plugin provides four deterministic scripts to engineer agent memory by optimizing what systems intentionally forget rather than just what they remember. It profiles construction vs. query costs per correct answer, picks optimal memory architectures against constraints, audits memory density, and enforces forgetting policies with blocking checks against designs that never forget on purpose.

Browser-automation skill for controlling Google's NotebookLM to read/query notebooks, add sources (URLs, text, files, YouTube links), and generate Studio outputs like Audio Overviews, Study Guides, Mind Maps, and Reports. Requires a browser automation environment (Claude Code CLI with computer-use or Claude Chrome Extension) and fails gracefully in non-automation contexts.

A Claude Code plugin that routes patent search requests to one of five specialized sub-use-cases (novelty, FTO, competitive landscape, acquisition diligence, or litigation prior-art) and executes a jurisdiction-aware, claims-focused search across Google Patents, Espacenet, USPTO, and optionally Lens.org. Output is an editable Word document with ranked closest art, CPC-class landscape analysis, family-resolved deduplication, FTO risk flags, and audit log—with mandatory legal disclaimers recommending attorney consultation before filing or licensing decisions.

This plugin bundles 13 production-ready product skills—including RICE prioritization, OKR cascading, persona generation, competitive teardown, landing page scaffolding, and SaaS boilerplate—each implemented as Python automation tools with zero dependencies. It equips product managers, strategists, UX researchers, and team leads with deterministic workflows orchestrated across four specialized agents and accessible via slash commands.

prompt-governance is a Claude Code plugin providing a skill for managing prompts at scale in production, including versioning, A/B testing, prompt registries, regression prevention, and evaluation pipelines. It enables teams to systematically control and optimize prompts without external dependencies.

Pulse is a multi-source recency research skill that gathers and synthesizes current discourse across Reddit, Hacker News, the open web, and optionally X/Twitter within a configurable time window (default 30 days). It enforces research integrity through intake questions, parallel execution with strict rate limits, three-count source tracking, and cross-platform pattern analysis before producing a citations-backed briefing.

A collection of 17 expert skills for HealthTech and MedTech regulatory affairs and quality management, covering ISO 13485 QMS, EU MDR 2017/745, FDA submissions, GDPR/DSGVO, ISO 27001, SOC 2, CAPA management, and audit compliance. The skills are organized across five strategic layers (leadership, core QM, risk/security, regulatory specialization, and audit/compliance) and work across Claude Code, Cursor, VS Code, and other AI agents without requiring external API keys.

A mid-conversation reflection skill that pauses execution and applies a 5-dimension analysis framework (macro perspective, gap analysis, reflective inquiry, bias check, contextual alignment) to reassess direction, assumptions, and bias when explicitly invoked or when conversation signals warrant deeper introspection. The skill delivers honest, specific reasoning—never manufacturing problems when the path is solid, and clearly stating evidence when redirection is needed, ending with a Continue/Pivot/Pause recommendation.

A hybrid research router that classifies any research request deterministically and either delegates to a specialist skill (pulse, grants, litreview, syllabus, patent, dossier) or runs its own plan-decompose-multi-source-search-synthesize workflow for general research. Always surfaces the routing decision so users can override, producing markdown briefings or .docx documents with full citations and audit logs.

A Claude Code plugin providing 4 specialized research-operations skills covering clinical-research, research-finance, market-research, and product-research workflows, plus an orchestrator skill that uses fork context to coordinate across domains. Each skill includes domain-specific Python tools (24 total), onboarding modules, and integration with an optional autoresearch evaluator for engineering workflows.

research-summarizer is an agent skill for Claude Code, Codex, and Gemini CLI that summarizes academic papers, compares web articles, and extracts citations into structured research briefs. It processes research documents without requiring external APIs or tools.

The /roast command convenes five independent adversarial reviewers (Critic, Champion, Analyst, Investigator, Customer) to attack a business idea from every angle in parallel, then a Judge synthesizes one GO/RESHAPE/KILL verdict with explicit confidence and a concrete 48-hour test. It replaces naive averaging with weighted veto gates on fatal flaws, broken logic, and missing buyer demand.

skill-doctor grades a Claude Code agent's setup by analyzing real conversation history from recent local sessions, scoring transcripts against fixed rubrics for efficiency and code quality, then proposing skill edits backed by evidence. All computation (scoring aggregation, validation gates, arithmetic) is deterministic and local; redaction, permissions, and strict evidence traceability ensure privacy and prevent suggestion fabrication.

This Claude Code skill enforces Google SRE Workbook discipline for defining SLOs with an SLO designer (requires all fields), multi-window burn-rate alert calculator (PromQL-shaped), and SLO reviewer that catches 7 common bugs like missing SLI definitions and unrealistic targets. It ships 4 reference docs on principles and SLI/error-budget design, 2 YAML asset templates, and a /slo-design slash command.

This plugin provides a Claude Code skill for Snowflake SQL development, including support for Dynamic Tables, Streams+Tasks, Cortex AI functions, and Snowpark Python. It includes a query helper script, 3 reference guides, and troubleshooting documentation for common Snowflake development scenarios.

This plugin provides a structured knowledge base extracted from "Spinning Up in Deep RL" by Joshua Achiam, covering RL fundamentals, model-free algorithm taxonomy, policy gradients, and six reference algorithms (VPG, TRPO, PPO, DDPG, TD3, SAC). Use it to apply Achiam's frameworks for debugging RL code, understanding algorithm derivations, and designing rigorous multi-seed experiments.

This plugin provides statistical hypothesis testing and A/B experiment analysis through 3 Python tools using only the standard library, covering Z-tests, t-tests, chi-square tests, effect size calculations, power analysis, and Wilson score confidence intervals. It enables users to run sample size calculations and determine confidence intervals for experimental design and validation.

A CEFR-leveled Swedish-learning mentor that probes your actual proficiency with two assessment questions, then builds a customized learning path across listening, reading, writing, and speaking skills. It recommends 3–6 vetted YouTube and podcast resources matched to your level, pulling from a maintained catalog including Peter SFI, Lätt Svenska med Oskar, Radio Sweden på lätt svenska, and others.

Generates a curated supplementary reading list from course syllabi by parsing topics and learning outcomes, then searching Consensus for recent peer-reviewed papers with audience-calibrated summaries and Bloom-aligned discussion questions. The skill structures topics into 6–12 sections with user checkpoints, runs targeted searches sequentially, and outputs a professionally formatted .docx with clickable paper links.

This plugin provides an agent skill for designing and implementing Terraform infrastructure-as-code patterns, covering module design, state management, provider configuration, and security hardening. It assists with mono-repo vs multi-repo decisions, workspace management, policy-as-code implementation, and CI/CD plan/apply workflows including drift detection.

This universal scraping plugin intelligently routes requests between Firecrawl and local Python extraction (using pandas, requests, beautifulsoup4) while tracking token budgets and quota limits. It supports free-tier usage with bring-your-own-key (BYOK) authentication for optional paid services.

This plugin provides a single skill for developing video content strategies, writing scripts, and optimizing channels across YouTube, TikTok, Reels, and Shorts. It helps repurpose long-form content into short-form formats using Claude's language capabilities without requiring external APIs.

This plugin helps VPs of Engineering analyze and optimize delivery throughput, hiring, team structure, and production discipline through DORA metrics, hiring funnel analysis, and org design frameworks. It provides strategic advisory across four core decisions: delivery throughput (DORA 4 metrics + bottleneck identification), engineering hiring (7-stage conversion + pipeline gaps), team structure (squad/tribe model with scaling triggers), and production discipline (on-call, incidents, deployment, SLOs).

This plugin walks the three-phase GTD weekly review ritual (GET CLEAR, GET CURRENT, GET CREATIVE) with deterministic scripts that inventory open loops, gate completion by enforcing all mandatory steps, and audit project health with a commitment score. It refuses to mark a review complete unless all five GET CURRENT steps are accounted for, turning the weekly review into a gating ritual rather than a guilt check.

Workflow-builder creates deterministic multi-agent workflow scripts (.js files) for Claude Code's Workflow tool, with an intake-first recommendation engine that infers topology when user requests are vague. It includes a validator enforcing pure-literal rules and no non-determinism, a scaffolder for 5 orchestration patterns (fan-out, pipeline, barrier, loop, judge-panel), and three reference guides with runnable templates.

This Claude Code plugin helps users create new agent skills with proper structure, progressive disclosure, and bundled resources by combining Matt Pocock's original 3-phase workflow (Gather → Draft → Review) with stdlib Python validation tools and in-depth reference documentation. It surfaces skill authoring through a cs-skill-author persona agent and /cs:write-a-skill slash command that operationalizes Matt's review checklist to catch common mistakes.

The youtube-full plugin extracts YouTube transcripts with timestamps, searches within channels, browses playlists, and monitors new uploads via TranscriptAPI. It covers complete YouTube data workflows with 100 free credits included (BYOK model).

Zero-Hallucination-Coder enforces a disciplined five-step pipeline (Discuss → Map → Decompose → Execute → Verify) that grounds code generation in verified structure and avoids invented APIs, assumed imports, or placeholder code. It applies YAGNI principles via a lazy-senior-dev ladder to delete unnecessary code first, making it suitable for high-stakes, complex, or multi-file work.

Taskmaster is a task management system designed for AI-driven development workflows within Cursor, providing 46 commands and 3 agents to organize work without overwhelming the editor. It operates entirely offline with no external dependencies or API keys required.

This plugin helps PMs and founders audit AI-generated code by reverse-engineering documentation from the codebase, then comparing documented intent against actual implementation to catch correctness, security, and performance defects. It produces a reviewer-ready shipping packet with architecture docs, permission flows, test coverage maps, and evidence-backed audit findings.

This plugin equips product managers with SQL query generation and cohort analysis skills to extract insights from user data. It provides three main capabilities: generating SQL queries from natural language descriptions, performing cohort analysis to identify retention patterns and engagement trends, and analyzing A/B test results with statistical rigor.

This plugin provides 16 product management skills for creating PRDs, OKRs, roadmaps, user stories, and running sprints, retrospectives, and risk analysis through 11 commands. It operates entirely offline with no external API dependencies, offering frameworks for stakeholder mapping, prioritization, pre-mortems, and test scenario generation.

This plugin equips product managers with structured frameworks for launching products, including GTM strategy, growth loops, competitive positioning, and customer segmentation. It provides six skills and three commands covering everything from identifying beachhead segments and ideal customer profiles to designing growth mechanisms and sales-ready battlecards.

This plugin provides product marketing and growth utilities including marketing idea generation, value proposition crafting, North Star metric definition, product naming, and competitive positioning analysis. It offers two main commands and five skills to help teams develop comprehensive go-to-market strategies and validate their product's core metrics.

This plugin provides market research skills for product managers, including tools to analyze competitors, build user personas, segment customers, and perform sentiment analysis on feedback. It offers three commands to conduct competitive analysis, analyze user feedback at scale, and run comprehensive user research with persona building and customer journey mapping.

This plugin provides 13 product discovery skills for PMs to run ideation, experiment design, assumption testing, feature prioritization, and customer interview workflows. It includes commands to brainstorm ideas, run full discovery cycles, manage metrics dashboards, and synthesize customer interviews without requiring any external APIs or tools.

This plugin provides 12 product strategy skills and 5 commands covering frameworks essential for product managers—including Business Model Canvas, Lean Canvas, SWOT, PESTLE, Porter's Five Forces, Ansoff Matrix, and value proposition design. It enables PMs to analyze markets, design pricing strategies, and develop comprehensive product strategies without requiring external APIs or services.

The pm-toolkit provides four core skills for product managers: resume review and tailoring against PM best practices, NDA drafting with jurisdiction-specific clauses, privacy policy generation covering compliance requirements, and grammar/flow checking with targeted fixes. It runs entirely locally without external API dependencies or third-party services.

This plugin teaches AI coding agents to create promptfoo eval suites with deterministic assertions, provider configurations, and testing best practices. It provides a skill for generating and validating evaluation frameworks that help ensure consistent prompt and model performance.

Compound Engineering is a 36-skill plugin that structures AI coding workflows around brainstorming, planning, building, and reviewing code while capturing learnings for future changes. It runs on 14 agent hosts including Claude Code, Cursor, and Codex, emphasizing leverage through planning and knowledge reuse rather than ceremony.

seo-cockpit is a mod companion for claude-seo that enforces budget limits on paid SEO API calls, logs spend, and surfaces a live terminal dashboard showing audit progress, Core Web Vitals, Search Console data, rankings, and spend in real time. It provides a spend guard that holds any API call it cannot validate, an audit band with live progress tracking, and an interactive visual cockpit with charts and tables fetched from free sources.

DeepEval is an open-source LLM evaluation framework that provides metrics and tools for unit testing AI applications, agents, RAG pipelines, and chatbots using LLM-as-a-judge evaluation and locally-running NLP models. The plugin adds three skills to enable DeepEval evaluations, tracing, datasets, and integration with the Confident AI observability platform directly into Claude Code workflows.

Browser Harness is a Claude Code skill that connects an LLM directly to Chrome or Chromium via Chrome DevTools Protocol, enabling coordinate-based clicks, screenshots, and Python helpers without selector hunting. The agent can drive your real browser or a Browser Use Cloud browser, and auto-generates helper code and domain-specific skills as it learns during execution.

/brag is an agent skill that transforms a project into a short, shareable launch video with music, motion, and share copy by analyzing the project and sending a focused brief to Hyperframes for rendering. It also includes /brag-slim, a lean variant for Claude Opus 5.5 that creates videos without requiring Hyperframes.

A comprehensive skill pack with 67 specialized skills covering 12 programming languages, backend and frontend frameworks, infrastructure, DevOps, security, and testing—designed for full-stack developers with context-aware activation and multi-skill workflows. Includes 9 workflow commands for project management integration with Atlassian tools and progressive disclosure architecture for faster loading.

Two skills for building on Cursor Origin, Cursor's code forge: one provides API guidance by fetching the live spec and listing rules for credentials, scopes, webhooks, paging, and errors; the other plans the port of an existing GitHub App by mapping its usage to the live Origin spec and writing a porting brief with what carries over, webhook fields, required scopes, tests, and team decisions. Both skills read the live spec at runtime and never reference endpoints from memory.

pstack is a Claude Code plugin with 2 agents and 54 skills that provides rigorous, parallelizable agent workflows for writing higher-quality code with less effort. It emphasizes a "go deep first" approach to code generation and problem-solving.

This agent skill converts complex terminal output like system architectures, diffs, and comparisons into styled, interactive HTML pages with dark/light themes and Mermaid diagrams instead of ASCII art. It provides commands for generating visual explanations, plans, and slides that open directly in your browser without requiring any build dependencies.

Claude Ads is a Claude-first paid-media operations plugin spanning 12 advertising platforms (Google Ads, Meta Ads, YouTube Ads, LinkedIn Ads, TikTok Ads, Microsoft Advertising, Apple Ads, Amazon Ads, Reddit Ads, Pinterest Ads, Snapchat Ads, and X Ads) with capabilities for audits, campaign planning, creative workflows, monitoring, and deterministic reports—all read-only by default with gated approval for live changes. The plugin composes 25 agents and 33 skills to deliver source-grounded analysis, evidence-backed optimization drafts, and versioned JSON outputs rendered as Markdown, HTML, and optional PDF.

An agent skill that generates immersive scroll-scrubbed "fly through the world" landing pages with AI-rendered isometric scenes and seamless camera flights. The skill interviews the user for content and branding, generates cohesive scene stills and continuous camera clips via GPT Image 2 and Seedance, then wires a portable vanilla-JS scrub engine that works with any web framework.

Harness is a Claude Code plugin that transforms a domain description into a coordinated agent team architecture and supporting skills by applying one of six pre-defined team patterns: Pipeline, Fan-out/Fan-in, Expert Pool, Producer-Reviewer, Supervisor, or Hierarchical Delegation. It automatically generates agent definitions and skills tailored to your domain through a structured six-phase workflow spanning domain analysis, architecture design, agent/skill generation, integration, and validation.

This Claude Code plugin uses the Jev model to intelligently compact conversation history by deciding which tool calls and results to keep, drop, or truncate—never summarizing or rewriting text. It preserves exact file paths, errors, and constraints while fitting the state within token limits through staged truncation.

This plugin audits a project's npm, PyPI, and Go dependencies for supply-chain risks by checking version-matched advisories (OSV), deprecated packages, abandoned repositories, publisher concentration, and install scripts. It generates a deterministic Markdown report with clear separation between measured data and model-provided narrative and remediation guidance.

A collection of 11 agent skills that help build great product interfaces by addressing typography, colors, layout, accessibility, UI polish, and product writing. Skills include reviews of existing interfaces, component variant generation, UI stress testing, and targeted improvements across design and accessibility domains.

A Claude Code plugin providing seven skills and two subagents for building, debugging, and auditing applications with the Atomic Agents Python framework. It includes creation workflows for schemas, agents, tools, and context providers; a diagnostic troubleshooter keyed to real error messages; a project scaffolder; and specialized subagents for codebase mapping and code review in isolated context.

This Claude Code plugin provides skills for optimizing and scaling Daft dataframe workloads, including UDF tuning and distributed processing guidance. It helps developers navigate Daft documentation and implement performance improvements for data engineering tasks.

This plugin scaffolds a local skill and sets up an iterated coding-agent workflow with accompanying prompt templates and memory file for persistent state. It enables developers to build multi-turn agentic coding loops without external dependencies.

This plugin interviews users about their codebase requirements, then designs and builds a locally-runnable agentic control loop with sensor, controller, and actuator components subject to real-world disturbances. The resulting system is deployed as scheduled coding-agent workflows integrated into the user's development pipeline.

This Claude Code plugin enhances instruction files by rewriting CLAUDE.md using <important if> blocks to improve how Claude follows your project guidelines. It's a single skill that helps clarify conditional instructions and boost instruction adherence.

This skill narrows React component prop types to match live code paths, helping developers enforce stricter type checking based on actual runtime behavior. It integrates with React codebases to analyze and refine prop type definitions for better type safety.

The rpi plugin guides you through researching a code change, working through its design, and implementing it with reviewable local documents. It combines 5 agents and 1 skill to help structure the entire change lifecycle from concept to implementation.

The show-me plugin generates visual explanations of topics through concise diagrams, code-shape sketches, and focused HTML artifacts. It provides a single skill that helps clarify concepts by rendering them in diagram and code visualization formats.

This Claude Code plugin creates or updates GitHub pull requests with concise visual outlines to help reviewers quickly understand changes. It operates as a single skill that generates structured PR descriptions summarizing the modifications.

<UNKNOWN>

Generates simple, visual HTML explanations of any topic using the /eli5 command, with large pictures and minimal text suitable for explaining concepts to a five-year-old. Requires no API keys or external tools, relying entirely on Claude's built-in capabilities.

This plugin generates a single interactive HTML page that organizes implementation plans as a hierarchical tree of claims, each supported by mockups, state machines, call stacks, schemas, or code snippets. Users can expand the tree progressively, answer embedded decisions, edit schemas inline, and respond with feedback that can be pasted back into Claude for refinement.

This plugin monitors conversation turns and suggests up to three contextual next prompts that users can select by pressing 1, 2, or 3 to draft in the input box. It uses a forked session with the model to generate relevant follow-up suggestions without automatically submitting them, and filters suggestions based on available commands and answer length.

AI media generation skill for QuickDesign CLI that generates images (Nano Banana, GPT Image) and videos (Sora 2, Kling, Seedance 2.0 R2V) with multi-segment voice continuity and reference composition support. Includes Smart Ad Creator for parallel concept generation, Spy Brands competitor library queries, and Brand DNA extraction from websites.

designlang extracts complete design systems from live websites—DTCG tokens, Tailwind/shadcn configs, Figma variables, motion tokens, and layout patterns—then grades consistency, generates brand-guidelines books, and measures visual fidelity against clones. Fourteen slash commands enable whole-site analysis, design-space positioning, head-to-head battles, theme recoloring, and shareable graded report cards with pixel-perfect diff scoring.

Treg is a curated catalog of 2,896 agent-friendly tools from multiple providers (SEO, scraping, data enrichment, content generation) accessible through a single token and base URL, with pay-per-call pricing starting at a cent per call and no provider signup required. Users can also register their own tools, APIs, and CLIs alongside the shared catalog, with all calls relayed server-side to keep credentials secure.

This plugin equips Claude Code with a skill that guides idiomatic and modern Go development practices, helping generate code that follows current Go conventions and best practices. It requires no external tools or API keys and works entirely within the Claude Code environment.

A portable Agent Skill providing expert guidance on SwiftUI state management, view composition, performance optimization, and modern SDK APIs including Liquid Glass adoption. The skill loads reference material on-demand to assist code review and SwiftUI development workflows without inflating context size.

This official NVIDIA plugin provides a curated skill for GPU acceleration, CUDA, AI agents, data loading, training, inference, robotics, Physical AI, Omniverse, and simulation workflows. The plugin is generated from NVIDIA's configuration and designed for delivery to official OpenAI and Anthropic marketplaces.

nateherk-design provides a skill for building scroll-driven interactive landing pages with comprehensive design control over typography, spacing, colour, depth, motion, and accessibility. The skill enables premium scroll-craft implementation without relying on animation libraries, using a full design floor approach.

OpenOutreach is a self-hosted, open-source B2B lead-finder CLI that discovers and qualifies prospects from licensed data based on your product description and target market, then optionally emails them directly. This Claude Code plugin skill teaches the agent when to invoke the OpenOutreach commands, how to parse the CSV output, and which operations incur credits.

terminal-browser is a Claude Code plugin that embeds a fully functional web browser directly into the Claude Code interface using the kitty graphics protocol, allowing users to preview websites and HTML documents. The plugin exposes an API for other Claude Code plugins to programmatically open URLs in the embedded browser, enabling custom integrations like slash commands for opening pull requests or design tools.

An Agent Skill that enables coding agents to write and execute custom Playwright browser automation scripts on the fly, from simple page tests to complex multi-step flows with real-time visibility. The skill provides dev-server detection, portable script execution, and smart cleanup while letting Claude autonomously decide when to apply it based on automation needs.

The mcp-apps plugin provides Claude Code skills for building MCP (Model Context Protocol) Apps with interactive user interfaces and UIs. It includes scaffolding, migration utilities for converting OpenAI Apps SDK projects to MCP, and tools for adding interactive views to MCP tools.

A collection of 12 Claude Code and Codex skills for LinkedIn marketing that draft posts, comments, and replies while helping remove AI tells and optimizing content for the platform. The skills cover post writing, comment drafting, profile optimization, content planning, thread monitoring, and engagement analytics—all requiring user approval before publishing.

This plugin provides experimental Expo skills for pre-release and evolving APIs that lack stability guarantees and may change between versions. The included `expo-migrate-module` skill assists with migrating Swift Expo modules from the Modules API 1.0 DSL to the 2.0 macro API.

A deterministic documentation and code consistency auditor that inventories artifacts across README, CLAUDE.md, CHANGELOG, CI workflows, and package manifests, then reports drift via 9 structural checks (version strings, CI commands, references, cross-doc facts) resolved by human-adjudicated authority per fact class in sot-map.yaml. Read-only and advisory-safe — deterministic Critical findings can block releases, but LLM-judged findings are advisory-only and never fix anything.

This skill automatically validates, generates, and transforms YAML files with intelligent error detection, schema inference, and format conversion capabilities. It handles Kubernetes manifests, Docker Compose configs, CI/CD pipelines, and general YAML work with linting, anchors/aliases mastery, and format conversions between YAML, JSON, TOML, and XML.

This plugin extends Claude Code with Google Vertex AI multimodal capabilities for marketing automation, enabling video processing up to 6 hours, audio generation via the Lyria model, and image creation with Gemini 2.0/2.5 and Imagen 4. It automates marketing campaign asset production including scripts, music, images, and social media content through the `/vertex-campaign` command and auto-activating skills.

This plugin provides Claude Code guidance and tooling for building, evaluating, and deploying Google ADK applications using Google's maintained `agents-cli` workflow across Agent Runtime, Cloud Run, GKE, and Gemini Enterprise. It includes a `/create-agent` scaffolding command and a repo-aware lifecycle skill that covers local development, evaluations, cloud deployment, and observability without embedding cached configuration or pricing data.

A collection of 18 production operator skills for health systems deploying Abridge clinical documentation, covering consent, clinician review, Epic handoff, access, security, reliability, and change control workflows. Each skill is grounded in public Abridge product documentation and clearly separates public facts from tenant-specific implementation details.

The access-control-auditor plugin audits access control implementations through a command and skill interface, enabling developers to review permission structures and security configurations in their codebase. It operates entirely offline without requiring external API keys or third-party services.

This Claude Code plugin performs automated accessibility compliance testing against WCAG 2.1/2.2, ARIA standards, and other accessibility frameworks via the /a11y-scan command. It validates keyboard navigation, screen reader compatibility, color contrast, and generates fix recommendations and test scenarios for identified accessibility issues.

The Adobe Skill Pack provides 30 evidence-backed Claude Code skills for operating Adobe Firefly, Photoshop API v2, PDF Services, I/O Events, App Builder, and enterprise identity with explicit auth boundaries, data custody contracts, and safe output requirements. Every skill includes first-party Adobe documentation maps, negative-path validation, and operates within predetermined approval boundaries without network access, credentials, or spend authorization.

This Claude Code plugin turns Notion into an AI agency dashboard where agents plan tasks collaboratively and execute work in parallel with intelligent model routing and dependency management. Users provide feedback and approve work through a single Notion board, with all results and task status linked back to the rows.

This Claude Code plugin automatically detects and loads AGENTS.md files alongside CLAUDE.md to provide agent-specific instructions and specialized behaviors. It operates through three redundant layers—automatic skill invocation, directory change detection, and a manual sync command—ensuring agent-specific rules are always available without requiring manual intervention.

A Claude Code plugin that scans local repositories for high-risk patterns (uncommitted changes, dangerous automation configs, shell-to-network chains, credential writes) and generates a color-coded preflight receipt (Green/Yellow/Red) before allowing AI agents to edit code. The bundled scanner runs entirely locally without API keys, network access, or external dependencies.

Agent Systems Toolkit provides model-neutral workflows for creating, validating, and production-upgrading portable Agent Skills, plugins, MCP integrations, and marketplace catalogs using specialist review roles and Beads-backed evidence. It includes three core skills—artifact-creator, artifact-validator, and production-upgrade—that operate without requiring a particular model or enforcing a single host-specific standard.

This Claude Code plugin analyzes git diffs and generates conventional commit messages automatically, offering three formatted options (concise, detailed, and comprehensive) for developers to choose from. It eliminates manual commit message writing by intelligently identifying change types, scopes, and breaking changes, then formatting everything according to Conventional Commits standards.

This plugin provides automated AI ethics and fairness validation through a command-line interface, implementing best practices and error handling for evaluating ML models. It requires Python 3.8+ and standard ML libraries, with no external API dependencies.

The AI Experiment Logger is a Claude Code plugin that tracks and analyzes AI experiments through both MCP tools and a web dashboard with local JSON storage. It provides structured logging, performance analytics, advanced filtering, and CSV export to help compare different AI tools and prompting strategies.

The AI/ML Engineering Pack is a collection of 12 specialized plugins organized across prompt engineering, LLM integration, RAG systems, and AI safety that work within Claude Code. It provides agents and commands to optimize prompts, generate production-ready templates, build LLM APIs, design RAG pipelines, and implement AI safety guardrails.

This plugin enables multi-agent orchestration using AI SDK v5, allowing specialized agents to hand off tasks, route requests, and coordinate workflows across OpenAI, Anthropic, Google, and Ollama providers. It includes 3 commands and 1 agent for setting up complex distributed AI systems with automatic coordination and optional local execution via Ollama.

Eighteen evidence-backed Claude Code skills for operating Alchemy Node, Data, Portfolio, NFT, Wallet, Notify, and Admin APIs across supported blockchain chains using current libraries like viem and @alchemy/wallet-apis instead of the archived alchemy-sdk. Each skill includes credential boundaries, error handling, security validation, and production readiness checks with zero network access or credential authorization during invocation.

This Claude Code plugin creates comprehensive alerting rules for performance monitoring, with features for multi-category alert creation, intelligent threshold setting, severity-based routing, and escalation policy configuration. It generates runbooks and provides alert testing procedures to help teams establish proactive monitoring strategies.

A production operator skill pack of 24 Claude Code skills for building, releasing, and operating Algolia search with coverage of search, indexing, credentials, events, migration, deployment, observability, incidents, privacy, cost, and performance. Skills are grounded in the Algolia JavaScript v5 contract and current first-party documentation without hard-coded pricing, invented quotas, or universal latency claims.

Anima-pack is a Claude Code plugin with 18 skills for automated design-to-code workflows, converting Figma designs and websites into React or HTML output using the Anima SDK 0.27.0. It includes standard workflows for generation and testing, plus pro skills for CI/CD integration, deployment, webhooks, and performance tuning with built-in security and review gates.

The anomaly-detection-system plugin detects anomalies and outliers in data through a command-line interface (/detect-anomaly) using Python and standard ML libraries. It provides automated workflows with performance optimization and error handling capabilities.

This Claude Code plugin creates Ansible playbooks for configuration management, featuring production-ready templates that follow best practices and support multi-platform deployments. The plugin includes one command and one skill to generate well-documented playbooks.

A 30-skill Claude Code plugin pack covering the Anthropic Messages API, tool use, streaming, Message Batches, prompt caching, and multi-model routing with production examples in Python and TypeScript. Skills span from installation and hello-world to advanced patterns like load scaling, reliability, incident runbooks, and enterprise RBAC.

Aomi bundles two agent skills: aomi-transact drives the Aomi CLI to chat, simulate, and sign on-chain transactions with account-abstraction execution across 40+ DeFi/perps/CEX/social apps on EVM mainnets and L2s, while aomi-build scaffolds new Aomi apps and Rust SDK crates from OpenAPI/Swagger specs. The bundle integrates with Claude Code, Cursor, Gemini CLI, OpenAI Codex, VS Code Copilot, and any Agent Skills-compatible AI tool.

This Claude Code plugin provides a command to build complete authentication systems supporting JWT, OAuth2, API keys, and role-based access control (RBAC). It operates as a self-contained tool with no external dependencies or paid services required.

This plugin implements batch API operations with job queues and progress tracking via a single command interface. It provides bulk processing capabilities for handling large-scale API requests efficiently.

This plugin implements caching strategies for APIs using Redis, HTTP headers, and CDN integration to reduce latency and improve performance. It provides a single command to set up and configure caching mechanisms across different layers of your application infrastructure.

This plugin generates API contracts for consumer-driven contract testing using Pact, enabling teams to define and validate API interactions between services. It provides a single command to create contract definitions that enforce API compatibility across consumers and providers.

Generates comprehensive API documentation from OpenAPI/Swagger specifications with interactive UIs and multiple export formats. Supports OpenAPI 3.0.3, Swagger UI integration, request/response examples, and automatic code sample generation.

This plugin provides standardized error handling for APIs with proper HTTP status codes and consistent response formats. It includes a command that implements error handling patterns through a reusable skill.

This plugin provides a command to implement event-driven APIs with support for message queues and event streaming. It includes a reusable skill for setting up the necessary infrastructure and patterns for event-based communication in APIs.

The api-fuzzer plugin performs fuzz testing on APIs by generating malformed inputs, edge cases, and security-focused payloads including SQL injection and XSS detection. It requires no external dependencies or API keys and operates through a simple `/fuzz-api` command.

This skill pack provides Claude Code with 18 production-grade skills for web scraping, Actor development, and data extraction using Apify, covering the apify-client library, Actor SDK, Crawlee crawlers, CLI, storage, webhooks, and deployment workflows. Skills span from installation and hello-world examples through debugging, rate limiting, security, production checklists, CI/CD integration, webhook management, and cost optimization on the Apify cloud platform.

This plugin generates API gateway configurations with routing, authentication, and rate limiting for microservice integration. It provides features including JWT authentication, load balancing, and circuit breaker patterns via a single command.

This plugin enables load testing of APIs using k6, Gatling, or Artillery—three popular open-source performance testing frameworks. Users invoke a single command (/run-load-test) to configure and execute load tests against their APIs for performance validation.

The api-migration-tool plugin assists with migrating APIs between versions while maintaining backward compatibility and ensuring smooth transitions. It provides a command-driven workflow for managing version migration tasks without external dependencies.

Creates mock API servers directly from OpenAPI specifications, enabling developers to test client code against realistic endpoints without hitting production infrastructure. Features include stateful responses, configurable latency, error simulation, and procedurally generated test data.

This Claude Code plugin provides a command to create monitoring dashboards that display API health status, performance metrics, and alert configurations. It requires no external APIs or third-party services and can be used standalone to visualize monitoring data through a single command.

This plugin implements rate limiting using token bucket, sliding window, and fixed window algorithms with optional Redis backend support for distributed rate limiting across multiple instances. It provides per-user and per-IP rate limit enforcement with configurable burst handling and standard rate limit response headers.

This plugin sets up structured logging for API requests with correlation IDs to track requests across distributed systems. It includes a setup command that configures log aggregation for better observability and debugging.

This plugin validates API responses against JSON Schema, OpenAPI specs, and custom rules through a single command interface. It contains one command and one skill for automated schema validation without requiring external API keys or services.

This Claude Code plugin validates API schemas against multiple standards including JSON Schema, Joi, Yup, and Zod. It provides a /validate-schemas command that developers can use to ensure their API definitions conform to the specified validation framework.

This plugin generates type-safe client SDKs from OpenAPI specifications across multiple programming languages. It provides a single command to streamline SDK creation from API specifications.

This Claude Code plugin scans APIs for OWASP API Security Top 10 vulnerabilities, including authentication, authorization, injection, and misconfiguration flaws. It provides remediation guidance to help developers address identified security issues.

This Claude Code plugin automatically generates comprehensive test suites for REST and GraphQL APIs, including authentication flows, CRUD operations, and contract validation against OpenAPI specs. It creates ready-to-run tests with fixtures, factories, and helpers that cover success paths, error scenarios, and security validation.

The API Throttling Manager plugin provides commands to configure and enforce dynamic rate limits and quota management for API calls. It includes one built-in skill to help implement throttling strategies with configurable rate-limit boundaries.

The API Versioning Manager plugin provides tools to handle API versions with backward compatibility, deprecation management, and migration strategies. It supports multiple versioning approaches, compatibility testing, and automated sunset headers to guide API consumers through version transitions.

This plugin generates comprehensive Application Performance Monitoring dashboards with golden signals monitoring and support for multiple platforms including Grafana, Datadog, and New Relic. It provides dashboard configuration generation, alert rule definitions, and visualization recommendations through a command-line interface.

Apollo Skill Pack provides 24 production-ready Claude Code skills for integrating with Apollo.io's B2B sales intelligence REST API, covering authentication, lead search, enrichment, email sequences, and deal management across dev, CI/CD, and enterprise deployment stages. The skills handle Apollo's non-standard conventions including x-api-key headers, separate search endpoints, credit models, and master vs. standard key types with correct error handling and real API patterns.

A Claude Code skill pack providing 18 integrated skills for AppFolio Stack REST API, enabling property management workflows including portfolio analysis, tenant onboarding, lease management, and production deployment patterns. All skills use direct HTTP requests with Basic Auth—no external SDK required—covering properties, tenants, units, leases, bills, and webhook event handling.

A 24-skill Claude Code plugin for automating Apple Notes via JXA, AppleScript, osascript, and Shortcuts on macOS, since Notes lacks a REST API. Skills span authentication, CRUD operations, export formats, CI/CD integration, and enterprise features like multi-account setup and data migration.

This Claude Code plugin analyzes application performance by profiling CPU usage, memory consumption, and execution time to identify bottlenecks and hotspots. It provides optimization recommendations based on comprehensive performance metrics and code-level analysis.

This plugin scans centralized and decentralized exchanges for arbitrage opportunities, including cross-chain and triangular paths, with profit calculation and flash loan execution support. It provides commands to find opportunities, calculate profits, and monitor real-time price spreads across markets.

This skill pack provides 18 governed workflows for integrating AssemblyAI's pre-recorded Speech-to-Text, Streaming v3, Speech Understanding, and LLM Gateway operations while enforcing security and privacy constraints. The skills cover installation, development loops, core transcription workflows, error handling, deployment, webhooks, performance tuning, and migration from deprecated LeMUR and Streaming v2 APIs.

The attio-pack provides 18 operator skills for building, releasing, and safely operating Attio CRM integrations, covering authentication, record/list management, API rate limits, webhooks, testing, deployment, and production reliability. Skills are grounded in Attio's current REST API, OAuth, and webhook contracts with built-in safeguards for tenant isolation, least privilege, and rollback capabilities.

The authentication-validator plugin checks authentication implementations against security best practices, validating password policies, JWT tokens, session management, MFA configuration, and account security across multiple auth methods. It provides actionable findings and grade-based reports to identify critical issues like weak hashing, missing MFA, and improper token expiration.

This Claude Code plugin connects to the Kobiton mobile testing platform, providing 5 skills to manage devices, upload apps, run automation sessions, and view test results directly from AI coding assistants. It integrates with Claude Code, GitHub Copilot CLI, Gemini CLI, Codex CLI, and Cursor, requiring only a Kobiton account and OAuth authentication.

This Claude Code plugin builds automated machine learning pipelines with integrated workflows, best practices, and performance optimization. It requires Python 3.8+ and standard ML libraries but has no external service dependencies.

The auto-scaling-configurator plugin provides a command-line interface to configure auto-scaling policies for applications and infrastructure across multiple platforms. It bundles production-ready configuration templates and best practices for common scaling scenarios.

Axiom is a collection of 13 production-ready TDD-tested Claude Code skills for iOS and Apple platform development, covering SwiftUI performance, debugging, UI testing, Liquid Glass design, and advanced topics like concurrency and data persistence. Skills are automatically suggested in Claude Code or can be invoked directly to assist with modern iOS 26+ development workflows.

The B12 Website Generator plugin automatically creates a professional website by collecting a business name and description from the user, then generating a pre-populated B12 signup link that drafts a production-ready site. Claude handles the entire process without requiring external API keys or tools.

This Claude Code plugin implements backup strategies for databases and applications using a single command-driven interface. It provides production-ready, multi-platform backup configurations following industry best practices.

A comprehensive 18-skill operator pack for BambooHR integrations covering OAuth/API key authentication, employee data pipelines, webhook governance, rate limiting, security, and production deployment patterns. Skills are grounded in BambooHR's official Python and PHP SDK repositories and include CI/CD integration, local development, diagnostics, and pre-release checklists.

This plugin provides a command-line interface to explore blockchain data including transactions, addresses, smart contracts, blocks, and tokens across multiple networks like Ethereum, Polygon, Arbitrum, and Optimism. It decodes transaction details, analyzes smart contract security patterns, displays portfolio holdings, and resolves ENS domains using public blockchain explorer APIs.

This plugin analyzes application performance across CPU, memory, I/O, and database layers to identify bottlenecks and prioritize them by severity. It provides root cause analysis, remediation strategies with code examples, and prevention recommendations to resolve performance issues.

This Claude Code plugin guides users through a 7-part brand strategy framework (Brand Truth, Audience Architecture, Cultural Context, Messaging Framework, Visual Language, Channel Strategy, and Measurement Framework) to build comprehensive brand foundations using the same methodology top agencies employ. The skill auto-activates on brand-related queries and produces structured deliverables including strategy documents, positioning statements, audience personas, messaging frameworks, and measurement dashboards.

This Claude Code plugin provides 18 governed skills for authorized Bright Data proxy, Browser API, and Web Scraper operations, covering authentication, core workflows, reliability, and production deployment. The skills separate credentials from REST keys, enforce policy and cost controls, and use offline fixtures by default with independently protected live operations.

Cross-browser compatibility testing using Playwright for local development and four cloud platforms (BrowserStack, Sauce Labs, LambdaTest, Kobiton) for real-device validation across Chrome, Firefox, Safari, and Edge. Includes screenshot comparison, CSS/JS API scanning, accessibility audits, and CI/CD integration examples.

This plugin analyzes and optimizes caching strategies through cache hit rate analysis, TTL optimization, and cache key design review. It provides recommendations for invalidation strategies, multi-level caching configurations, and memory usage optimization.

This Claude Code plugin analyzes DXF drawings through a single skill that drives the `cad-analyze` CLI, providing deterministic results for code compliance (ADA/IBC), drawing QA, quantity takeoff, summaries, RFIs, and room/zone detection without requiring any LLM or API keys. Users can ask Claude to check floor plans for compliance, generate health reports, or perform other CAD analysis tasks directly in Claude Code.

This plugin automates calendar-based workflows by generating meeting prep materials with agendas and participant backgrounds, and creating standup notes by synthesizing calendar events with related GitHub and Slack activity. It integrates with Claude's calendar Skills to enable commands like "prep for today's meetings" and "create standup notes from yesterday."

This Claude Code plugin provides 30 governed workflow skills for building, integrating, diagnosing, securing, deploying, and migrating Canva Connect backends using OAuth 2.0 PKCE, design operations, and asynchronous jobs. The pack covers authentication, authorization, observability, incident response, and production controls with evidence-backed patterns drawn from Canva's official documentation.