skills
Plugin for analysts, founders, investors and consultants that sizes markets bottom-up from Economic Mind reports, builds company financial snapshots, and benchmarks competitors on hard financials. Works with the Economic Mind connector.
4,248 plugins found in public marketplaces and GitHub repositories, indexed and searchable. Plugin skills also install in Codex, Cursor and 4 more agents. In Claude Code, the 99 that get Auto-invokedWhat is this?This plugin ships a FLOW.md router the engine fires, so the matching skill runs itself. No slash command to remember.Learn how → fire as you prompt.
Plugin for analysts, founders, investors and consultants that sizes markets bottom-up from Economic Mind reports, builds company financial snapshots, and benchmarks competitors on hard financials. Works with the Economic Mind connector.
A Claude Code skill that compresses large text, log, or document files with LLMLingua before processing them, so long pastes and transcripts use fewer tokens.
Skills and slash commands for LinkedIn growth that audit your profile, draft posts, carousels and comments, analyze performance, and remove AI-sounding wording from text. Open source under the MIT license.
A skill for Claude Code that makes the agent state a prediction about the result before changing code to fix behavior from an uncertain cause, so the next action tests the hypothesis.
MUNDO - THE EMPEROR. Complete AI orchestration system with 1208 skills, 25 capability modules, self-evolving, collective consciousness. GitHub Actions 24/7 automation.
Push it to GitHub and send us the link. We read it, then it goes live.
Lets the Claude Code agents on your machine chat with each other in channels or direct messages, with threaded replies and unread notices. Data is stored locally in SQLite, so no server is needed.
Before drafting a memo, PRD, decision doc or business review, it asks the questions that expose the real problem and locks the key decisions. Then it helps write the document and pressure-tests it before it ships, so the first draft holds up under scrutiny.
Enforces hard rules for coding agents through hooks, with an LLM judge that checks agent actions against your rules. Configured globally or per project, with a web config UI and an OpenCode adapter.
Gives an AI coding assistant shared working rules and quality checks, so its output stays consistent across a project. Adds guidance for backend, frontend, FastAPI and Rust work, and helps you plan endpoints and data models first.
Lets you leave anchored review comments on a repo's docs, then have your AI agent apply the feedback, resolve the threads, and log what changed and why. Includes setup and an authoring guide for the doc rendering palette.
Tailors a resume to a job description with honest fit analysis and keyword matching, exports it as an ATS-safe Word and PDF file, and saves your career profile for reuse on later applications.
Lets you queue coding tasks as GitHub Issues in a private repo you own, so named AI agent workers can claim, run and record each one while you stop babysitting terminals. Setup, status checks and worker runs are handled by its bundled skills.
A product marketing advisor exposed as an MCP server, drawing on a corpus of PMM books, podcasts and practitioner writing to help with positioning, messaging and battlecards.
A Claude Code plugin of skills, subagents and slash commands for academic work. It helps write thesis and paper sections, maintain BibTeX files, verify citations, expand the literature through citation graphs, and report benchmarks fairly.
Your agents have a theory. This skill makes them ask, “What would prove me wrong?”
FlowForge n8n is a local-first assurance layer for n8n workflow JSON: deterministic validation against bundled node metadata, static security/reliability/cost analysis, regression testing, snapshots, self-healing, semantic review, offline CI, and MCP tools.
Ultan gives Claude Code a persistent memory library of your preferences, coding conventions and past sessions, so it can recall them across projects. Includes skills to search the library and surface connections between entries.
Generates a graded comprehension quiz from the current Git branch diff, delivered in the terminal or through Google Forms. Useful for checking that a developer understands the changes before review or merge.
A skill for Claude Code and Codex that turns a website URL, brand name, or screenshots into a brand system with design tokens, a browsable brand book, and a landing page in the brand's layout style.
A visual team builder and live execution graph for autonomous Claude Code agent crews. It generates a plan as a dependency graph, runs the agents for hours, and lets you watch progress in the browser, with skills for research, cleanup of stale runs, and self-diagnostics.
Agent skills for social media managers and agencies that work with Planable, covering calendar audits, post performance patterns, cross-client metrics, and campaigns built from SE Ranking AI-search gaps.
A Claude Code plugin that checks changes to Shopify embedded apps by driving the real, authenticated Shopify admin in your own browser. After the agent edits app code, it verifies the result against the plan and repeats until the check passes.
Applies one developer's global Claude Code rules automatically in every session, enforced by hook scripts. Also adds skills for architecture diagrams, Jira ticket delivery, and responding to pull request review feedback.
Skill files for Claude that help with Couchbase, covering Analytics administration, Capella cloud management, cluster configuration, data-source links, and Analytics MCP setup.
A Claude Code plugin that reviews your project with a committee of AI personas chosen for its phase. A profiler reads the project and interviews you, then picks which reviewers take part.
A pre-ship audit for apps built with AI coding tools. It checks security, reliability, performance, AI integration, privacy and launch readiness, flags vulnerabilities, and gives the app a readiness score.
A Claude Code plugin that stops the coding agent from stopping at partial work. It carries a task through to the follow-up steps, such as adding tests and wiring up CI, until it is fully done.
A guide for engineers who want Claude Code to handle real work, from simple prompting to coordinated autonomous development. Also ships a deploy skill for releasing, rolling back and checking service health.
Wikispec is a spec-driven development plugin for Claude Code that keeps a Karpathy-style LLM wiki alongside OpenSpec changes. It provides skills to propose, explore, implement, and archive changes.
A decision framework for content operators that judges whether a topic is worth writing and in what priority, screening out weakly related or low-converting topics. Packaged as a skill for Claude Code and Cursor.
Adversarially reviews a drafted plan, spec or design doc before you execute it, using several AI coding CLIs you already have logged in, so flaws that would cause rework surface early. No API keys needed.
A Claude Code plugin that automates project documentation: architecture decision records, changelog entries, CLAUDE.md upkeep, dated development session logs, and a single navigation index linking your docs.
Single-CVE applicability triage for security teams. Ask whether a specific CVE affects your environment and get a reasoned verdict that cites each condition to its source and names every assumption made.
AI SAST framework for web & mobile apps, shipped as a Claude Code plugin. Agents read your source code and produce a validated, evidence-backed vulnerability report, no running the app, no network requests.
Claude Code plugins that hand tasks to external AI command-line tools, Codex, Gemini, and Grok, with an optional umbrella plugin that dispatches across providers. Useful for developers who want Claude Code to delegate planning, research, diagnosis, or review to these models.
Adds mojiemoji image stamps from mojiemoji.jozo.beer to Japanese GitHub Markdown such as issues, pull requests, reviews, replies, and release notes, emphasizing single words inline. Also manages stamp settings and catalog updates.
Adds a conductor skill that guides Claude Code through coding tasks in a git repo, such as implementing features, fixing bugs, refactoring, and debugging, with session guardrails and a code review step.
A Claude Code plugin with skills for reviewing software architecture, checking engineering estimates, deciding how to add features to an existing product, reviewing a whole codebase for readiness, and preparing for enterprise security questionnaires.
A Claude Code skill for building static sites with the Astro framework. It helps create Astro components and pages, set up content collections, optimize images, configure SEO, and deploy static builds.
A git pre-push hook that blocks a push until the developer answers a short multiple-choice quiz about the outgoing diff. An exam skill runs the quiz when a push is blocked or on request.
A project lifecycle workstation that coordinates agents and skills across a project's workflow, from setting up a new or existing project through requirements analysis, task breakdown, and scaffolding to delivery.
A Claude Code plugin with specialist agents, skills for agent creation and handoff, and hooks. It adds a memory that outlasts the session and a consent gate before any action that changes state.
Builds an AI clone of a public author's reasoning from their public content, stored in a single SQLite knowledge base. The clone answers in the author's frameworks and cites the source passages.
Give Claude a PDF, screenshot, or job posting and ask it to build a resume, tailor one to a role, or adjust its layout. It returns a print-ready HTML file you can export to PDF.
Tracks provenance, confidence and mock-taint through JavaScript and Python code so uncertain data does not appear as a confident result. Adds review-time checks, a GitHub Action, and skills that audit, bootstrap and remediate AI-assisted codebases.
A Claude skill that digs through your old Claude and ChatGPT chat history and saves forgotten to-dos, ideas, plans and drafts as local markdown notes organized by project. Runs locally, with no account needed.
A library of agent skills for Claude Code covering brainstorming, parallel subagent dispatch, executing plans, finishing git branches, and handling code review feedback.
A Claude Code plugin that scaffolds a new project from an empty folder, such as a SaaS, web app, API, or CLI, and hardens existing repositories to a standard. It also provides skills for commits, pull requests, and safe parallel editing.
Adds graph-based memory to Claude Code that records past interactions and learns recurring patterns across sessions. Runs as hooks and an MCP server.
Routes Claude Code work to the cheapest model tier that still holds quality, so the main session plans while cheaper models carry out routine tasks and lower token costs.
A Claude Code plugin for spec-driven development: it helps you write requirements as a specification first, implement code from that spec, and generate minimal documentation, all from the terminal.
Set the time your Claude 5-hour usage window starts each day, so resets land when you want them, not whenever you happen to start working. This creates a tiny cloud routine on your account that pings Claude 4×/day (every 5 hours) with a no-op message.
Shows the track playing on your Mac in the Claude Code statusline, and lets you pause, skip, check the song or view its cover art from the chat. It works with whatever app is playing, with no accounts or API keys.
Architecture Decision Record orchestrator for Claude Code. Given a feature request, issue, or problem statement, it explores the codebase, compares several architectural approaches, and writes the decision record. A companion skill structures vague problems into an intake document first.
Downloads paper PDFs from a DOI, or from a title resolved to a DOI first, trying Unpaywall, arXiv, bioRxiv and medRxiv, PubMed Central, Semantic Scholar, and Sci-Hub mirrors in turn. Supports batch downloads.
A Claude Code plugin for frontend UI work that generates code following a team's existing design system and component library. It also includes skills for turning specs into GitHub issues, picking up tasks, rebasing feature branches and shipping releases.
Drafts monthly or quarterly investor updates for a startup, structured the way investors tend to find useful. Use it when a founder asks to write, prepare, or send an investor update.
Stores your API keys in the OS keychain so Claude Code can use them without seeing the actual values. The model writes a placeholder, and the real key is substituted when the command runs.
A skill that makes Claude Code verify end-to-end outcomes across components, environments, and external systems before it treats a task as complete, labeling the evidence it gathers.
A skill for coding tasks that reads the codebase's conventions and sizes each change's risk before writing code, then verifies the result. It checks changes against the repo's own standards before they are accepted.
Builds and maintains a branded, interactive user's manual for a codebase as a single HTML file. Readers can annotate headings with ideas, feedback, or bugs, and a companion skill turns those notes into developer tickets.
A skill that draws hand-drawn-style diagrams from a description and renders them to PNG, covering flowcharts, architecture, sequence and ER diagrams, graphs, and mind maps.
Audits a web page or site for loading speed and responsiveness using Core Web Vitals (LCP, CLS, INP, TBT), producing a structured report with suggested fixes for images, fonts and JavaScript.
Checks an agent's work before you ship it, using independent reviewers who grade it against your standards, and pretests marketing copy on a simulated audience. Also gives sales advice on pricing and persuasion, and a file-based memory that carries context across sessions.
A Claude Code plugin with skills for React Native apps: fixing failing iOS and Android builds, safely upgrading Expo SDK and React Native, decoding Hermes crashes, handling deep links, and running EAS builds and store releases.
Transforms and exports data with DuckDB SQL. It reads CSV, Parquet, JSON, Excel and database files, applies joins, aggregations and pivots, and can write the results to files.
Takes a ticket from your tracker through an isolated worktree to an open pull request, then handles merging and branch cleanup. Includes batch dispatch of several tickets overnight and a spec step before dispatch.
Gives go-to-market teams Claude Code tools for enriching CRM records, coaching sales calls, preparing for discovery calls and building SaaS cohort and finance reports, so routine revenue work runs without paid software.
Flags documentation that still describes code you just changed, then repairs the affected docs against the current source. Runs as a post-edit hook or through a /stale-docs audit command.
A language and domain agnostic engineering skill for AI coding agents such as Claude Code and Cursor, guiding them to build services, parsers, CLIs and data pipelines that hold up under real use.
A Claude skill that grades, pressure-tests, and writes creative and campaign briefs and social hooks against a strategic standard (Idea, Insight, Objective, Benefit), so briefs move the people who make the work.
Helps you close out a session without losing loose ends, re-enter a long-running task after a break, and build work straight from a ticket, plan or description. Install once and the helpers load in every session.
A local-first personal voice layer for AI agents. Bring your voice to any agent. Adapt to any audience without losing yourself. Humanize AI-generated drafts by rewriting them in your own calibrated voice: PR comments, Slack replies, docs, RFCs, and emails.
A Rust-based codebase search service for coding agents that finds keywords across many files and checks whether a match sits inside a specific syntax node such as a function, class, or struct. Ships as a skill and an MCP server.
A skill that makes Claude Code check Git history before changing unusual or defensive code, mature APIs, dependencies, or architecture, to uncover hidden constraints behind existing code.
Lets Claude Code inspect a remote server read-only, reading logs, configs, process state, disk usage and journald without manual copy-pasting. A small Go binary runs as a diagnostic server on the remote host.
A Claude Code plugin for vibe coding that keeps project docs (CLAUDE.md, MODULE.md, handover.md) up to date, sets up lint and type-check guards, analyzes Python and JavaScript/TypeScript repo structure, and enforces root-cause debugging to catch AI coding mistakes.
A skill for building Zalo Mini Apps, lightweight React web apps that run inside the Zalo super-app. It covers setup, the zmp-cli and devtools, ZaUI components, zmp-sdk APIs, permissions, and phone number access.
Drive an Android emulator from your LLM. Build, install, launch, and test a feature end-to-end, get a unified logcat-and-actions timeline back.
A skill that guides beginners and vibecoders from a raw project idea to a secured, GitHub-ready repository, covering ideation through setup so common early mistakes are avoided.
Generates ready-to-use design artifacts such as blog hero images, business cards, carousels, Instagram posts and landing pages as pixel-exact HTML files. Also includes export tools for PDF and HTML-to-PNG.
Hold a natural spoken conversation with a coding agent, Claude Code or OpenAI's Codex, while it does real agentic work. Your Mac runs everything, the audio pipeline, the sessions, the state.
A planning plugin for Claude Code that guides an agent through questions, research, and review before implementation, then writes the result as a reviewed Markdown plan. Skills scale the plan to the size of the change and add diagnosis or reference research first.
Shows what an AI coding agent's diff changes in behavior, flow, contracts, and architecture, with each claim tied to the real code. Sets up a baseline and renders a shift card after each change.
An autonomous loop skill that takes any goal, such as coding, research, content, or data work, asks a few questions, then assembles a team of agents that discovers resources, implements, verifies, and iterates until the goal is met.
Audits what your Claude Code agent did against what it said, reads session logs and git, cites evidence, calls no model.
A Claude Code plugin for smart contract development on Solana with Anchor and on EVM chains with Foundry. It covers scaffolding, building and testing, auditing, and deploying contracts.
Gives your AI coding agent ready-made instructions for scaffolding a project, comparing product prices across stores, writing SQLAlchemy database code, and mapping how a codebase fits together. Once installed, the agent applies them on its own when a request matches.
Gives Claude Code a way to check its own work against a rubric and real test results, and to remember lessons across sessions. Instead of growing the prompt each time, a failed gate or bug is turned into a saved rule it reuses later.
The everything-Flow for building, shipping, and growing software, 40 hand-picked skills from 7 open-source suites, behind one FLOW.md router that fires the right skill on intent (and disambiguates the overlaps), with mandatory routing that survives context.
© 2026 Flowy
Built for Claude Code · Not affiliated with Anthropic
Logos are trademarks of their owners.