npm plugins

Author, edit, validate, and test Microsoft Copilot Studio agents using YAML files. Clone agents from the cloud, modify topics, actions, knowledge sources, and global variables, then push changes back. Run evaluations and batch tests against draft or published agents.

Run a complete product lifecycle from ideation to deployment: brainstorm specs, implement with TDD, review code, run security/SEO/performance audits, deploy with canary monitoring, and track post-ship metrics. Includes currency guard for up-to-date API/docs knowledge.

Manage Nx monorepos through Claude: run generators, execute tasks, explore workspace structure, fix dependency resolution errors, and drive autonomous CI fix loops that retrieve failures, analyze errors, apply fixes, and iterate until green.

Guides a full development lifecycle from vision and feature ideation through structured implementation, automated code quality audits, expert review, TDD cycles, safe refactoring, and production readiness checks — all within a single CLI-driven workflow.

Create and review software architecture documentation using AI specialist agents that assess design decisions, security, performance, maintainability, and simplicity, then record decisions as ADRs while enforcing YAGNI principles to prevent over-engineering.

Manage Skyboard kanban boards on AT Protocol from terminal CLI: join boards by URI, list boards/columns/cards, create/edit/move/delete/comment cards, view details, login/logout with OAuth.

Generate AI images and videos from text prompts via Leonardo.Ai CLI in your terminal. Specify models, dimensions, enhancements; poll job status, download results to files/directories; list models and check credits.

Automatically detect repeated multi-step workflow patterns in Claude Code sessions, generate reusable skills, and load them dynamically mid-session to automate common tasks like search-and-fix or find-and-update.

Monitor and debug Claude Code agent sessions with structured logs, knowledge tracking, and an event store. Query context-inject hooks, view MCP tool calls, search a knowledge base, and manage installation status from the CLI.

Identify emerging developer trends before they hit mainstream by scanning GitHub trending repos, package registry download spikes (PyPI, NPM), and tech Twitter conversations.

Run a complete PDCA (Plan-Do-Check-Act) knowledge work OS inside Claude Code: decompose tasks into parallel worktrees, research blocked URLs via adaptive fetch chains, review drafts with multi-perspective agents, and evolve prompts through benchmark-driven optimization loops.

Generate full architectural analysis reports mapping project components, dependencies, relationships, risks, bottlenecks, and infrastructure; audit dependencies across languages and package managers for vulnerabilities, outdated libraries, CVEs, and licenses; deeply analyze individual components for internal structure, business rules, design patterns, coupling, security, and technical debt. Produces timestamped Markdown reports saved to /docs/.

Audits, refines, and builds polished frontend UI using real-world design patterns, brand-appropriate color palettes, and anti-pattern detection — makes AI-generated visuals less generic.

Build Pi coding agents with grounded skills: deploy vLLM GPU pods on DataCrunch/RunPod, integrate via RPC/Node SDK in TypeScript apps, embed Lit web UIs and TUI components, author extensions/packages, customize prompts/behaviors, manage CLI workspaces with tmux, and deploy Dockerized Slack bots.

Launch and manage spec-driven marketing campaigns for developer products with automated positioning enforcement, quality gates, SEO audits, brand-voice content generation, and Vercel landing page deployment. Handles the full campaign lifecycle from brief to archive, including drift detection and learning extraction.

Build and customize AI chat interfaces in React/Next.js apps using assistant-ui library. Scaffold projects via CLI with templates, customize UI primitives and state with hooks, add multi-thread management and cloud persistence with auth, integrate LLM tools and streaming backends, and migrate packages to latest versions.

Automate Indian stock market research workflows: fetch concall transcripts from screener.in/BSE/NSE, extract variant perception scorecards and growth triggers via AI analysis, generate equity deep-dive reports with NotebookLM, and publish professional PDFs/HTML to Netlify.

Manage Business Central AL development with an engineering manager agent that spawns parallel specialist teams for requirements gathering, competitive solution design, multi-developer implementation, 4-specialist code review, and parallel test generation, plus a lightweight /fix workflow and documentation generation from project artifacts.

Standardize AI-generated projects with Ralph autonomous loop, persistent memory, and quality gates. Enforce Git workflow standards, code quality gates via pre-commit/pre-push hooks, and spec-driven task management with OpenSpec format. Manage skills and tasks through CLI commands and an MCP server for Cursor or Claude Code integration.

Automates git workflows, GitHub operations, security audits, and code quality enforcement for JavaScript/TypeScript projects, including conventional commits, branch management, PR creation, npm supply chain hardening, GitHub Actions audits, file naming and project structure audits, Tailwind CSS checks, and testing with Vitest.

Automate BMad Method project workflows including code audits, documentation, release management, issue/PR triage, adversarial reviews, LLM prompt auditing, root cause analysis, and skill publishing to Gemini/ChatGPT.

Prepare GNOME Shell extensions for extensions.gnome.org submission: scaffold projects with compliant structure, run automated linting and compliance checks, simulate EGO reviewer audits with rejection scoring, perform code reviews and field tests, and generate readiness reports.

Automate a complete build-test-fix loop: run precommit checks, test every commit in a branch, iteratively fix failures via fixup commits and autosquash rebasing, and block session end until code is reviewed.

Execute Hamster Studio briefs from Claude Code: schedule tasks in parallel waves, run diff-aware QA, perform paranoid code review, and generate engineering retrospectives.

Manage the full Dynamics 365 / Dataverse lifecycle: explore metadata, CRUD records, debug plugins, manage solutions and ALM, audit security, and configure web resources, webhooks, and service bus endpoints — all from Claude.

Migrate Next.js projects to vinext, a Vite-based reimplementation, by scanning compatibility, replacing packages via npm/yarn/pnpm, generating Vite configs, converting to ESM, and configuring deployments for Cloudflare Workers or Nitro.

Configure FeatureDrop toolkit in React apps via Claude Code to add changelogs, onboarding tours, checklists, badges, hotspots, feedback widgets, and surveys optimized for Tailwind or shadcn-ui styling, accelerating product adoption setup.

Author structured spells in a DeFi DSL using Grimoire CLI to fetch real-time market data from Aave V3, Uniswap, Morpho Blue, Pendle, Polymarket, Hyperliquid, and Across, then compile, validate, simulate, and cast them for deterministic agentic execution with explicit AI boundaries.

Develop and test Ignition SCADA projects with API references for Jython scripts and expression language, automated linting, and scaffolding for Jython unit tests and Playwright end-to-end tests for Perspective views.

Launch a browser-based GUI to assign Zelda BotW and TotK sound effects to Claude Code events like session starts, notifications, tool uses, and subagent actions. Preview sounds, save overrides instantly, and hear them play automatically via Node.js hooks during interactions.

Automate Brazilian legal workflows: analyze civil/criminal processes, draft judicial decisions, calculate CPC/CPP deadlines, download PJe case files as PDFs or separate documents, extract text from scanned legal PDFs via OCR, search unified CJF jurisprudence, and generate executive procedural reports in DOCX/PDF with institutional styling.

Enforce supply-chain risk policies on npm and pnpm package installations, and block git commits with unacceptable messages.

Build and extend FastMCP MCP servers in TypeScript through interactive code generation for custom tools with Zod schemas, static resources, authentication setups including API keys and OAuth, plus patterns for logging, streaming, and merge-friendly import sorting to cut Git conflicts.

Run cheminformatics workflows on SMILES, SMIRKS, and InChI notations via a RDKit-powered WASM CLI: validate structures, repair errors, convert formats, and perform analysis, with structured JSON output for scripting and agent integration. Installs via npm for fast subprocess calls from Node or Bash.

Run structured workflows and CLI commands to diagnose, auto-fix, manually intervene, and verify markdownlint errors (e.g., MD029 ordered lists) in .md files or folders. Reports total scanned files, error types, counts, file names, and line numbers using markdownlint-cli2.

Upload markdown files to Arweave for permanent decentralized storage with automatic provenance tagging including IPFS CID, author name, and app details. Receive ViewBlock transaction URL, direct Arweave link, and IPFS CID for verification and local logging via CLI command or skill invocation.

Securely install, scan, audit, and monitor packages across 14 ecosystems (npm, yarn, pnpm, bun, pip, uv, poetry, cargo, go, gem, composer, nuget, hex, pub) with CVE detection via OSV.dev, auto-fixes, severity triage, dependency audits, and continuous email alerts using INFYNON CLI and agent.

Orchestrate multi-agent teams in isolated Git worktrees for Claude Code to prototype divergent designs, debate trade-offs, perform premortem risk analysis, stress-test vulnerabilities, review code quality/security, and automate full PRD builds from spec decomposition to verified merging.

Scaffold new or migrate existing monday.com code apps with React/Vite frontend and Express/Node backend to serverless/CDN deployment. Run local dev servers with Vite, Express, and Docker MongoDB. Manage production with multi-region deploys, env vars/secrets, cron jobs, alerts, security scans, logs, and status checks via Monday API.

Orchestrate multi-agent autonomous development workflows: decompose features into testable stories, run parallel SDLC pipelines with isolated git worktrees, fix build/lint/TypeScript errors, execute TDD cycles, and inspect swarm state—all while enforcing tool-use policies and persisting session context via reflective memories.

Automate browser testing, social media scraping, CRM management, email triage, meeting follow-ups, and content generation across 1,000+ apps using AI agents and composable skills within a secure, company-configured environment.

Orchestrate the full product development lifecycle with phase-aware multi-agent skills: design debate, parallel implementation, cross-specialty review, security scanning, deployment gating, and production incident response — all recovered from workflow state on restart.

Validate .claude plugin folder structures and rules YAML on writes/edits/bash ops, enforce npm/yarn/pnpm via lockfiles, inject CLAUDE.md project context for discovery, auto-mkdirs, run TSX via npx, and guide Feature-Sliced Design implementation with custom views layer in Next.js apps.

Run end-to-end AI-driven feature development workflows in Claude Code: bootstrap ideas into PRDs, research codebases, draft plans and tasklists, implement iteratively via git edits and bounded scopes, review changes, execute QA tests across Python/pytest/npm/go/maven projects, with safety hooks gating prompts/files/executions.

Generate complete Docus-based documentation sites with search, dark mode, MCP server, and llms.txt integration for Nuxt projects supporting monorepos and various package managers. Review Markdown docs for quality, clarity, SEO, technical correctness, and Docus/Nuxt Content optimizations with prioritized recommendations.

Orchestrate structured AI debates with Toulmin arguments: create debate motions and setups, run multi-turn exchanges between proposition/opposition agents judged by an impartial scorer, and generate reports with analysis, verdicts, and argument graph visualizations.

Orchestrate Model Context Protocol (MCP) server lifecycles by installing, removing, listing, and updating tools for GitHub search, SQL queries, and Figma access. Resolve capability gaps through troubleshooting and discover new integrations to extend AI agent capabilities.

Index JS/TS dependency docs and source code locally, then search them with semantic retrieval and reranking — all offline via Docker containers.

Orchestrates a complete feature development lifecycle from codebase analysis and TDD test strategy through implementation, multi-agent code review (quality, security, performance, bugs, conventions), UI verification, and PR creation with resume capability for interrupted workflows.