jsapi-skills
This repository provides AI assistant Skills for Baidu Map JSAPI. Use them with Claude, Cursor, or other clients that support skills so the AI can reference accurate API docs and examples when you work with Baidu Map.
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.
This repository provides AI assistant Skills for Baidu Map JSAPI. Use them with Claude, Cursor, or other clients that support skills so the AI can reference accurate API docs and examples when you work with Baidu Map.
Finds prospects who actually reply to your outreach, drafts answers to replies for your approval, keeps warm leads from going cold, and shows which objections come up most. Works on its own or with La Growth Machine.
A Claude skill that converts HTML, CSS, JS, screenshots, PDFs, Figma exports or design mockups into WordPress Elementor template kits. Built for designers and developers who build WordPress sites.
Runs AI code review on pull requests and merge requests, posting findings directly to GitHub, GitLab, or Bitbucket instead of the terminal. It learns each repo's conventions and can apply fixes as a single commit with thread replies.
Push it to GitHub and send us the link. We read it, then it goes live.
Use Gemini from inside Claude Code for code reviews or to delegate tasks to Gemini. Based on openai/codex-plugin-cc, adapted for the Gemini CLI. Also adds /gemini:task for direct task delegation.
Compresses large files and noisy tool outputs such as command results, diffs, logs and search results before they enter Claude Code's context, reducing token usage.
A Claude Code plugin for full-stack development, offering modes for building, fixing, reviewing and deploying code, plus layered security hooks and token budget monitoring to manage context use.
Public, source-cited Obsidian brain for Claude Fable 5, Claude Mythos 5, Project Glasswing, Claude Code, and current agent-operations practice. This is a public research repository.
Split apart large applications and develop faster with microfrontends. This package contains library code to be used with Vercel's microfrontends support. Improved developer velocity: Build and test just the application that you are developing.
A Claude Code plugin with skills for hand-drawn architecture diagrams, code and architecture review guides, frontend design-system references, team knowledge wiki capture, and screen specification documents from PRDs.
Gives your AI assistant the OutSystems tools and working rules in one install, so it can inspect and build OutSystems apps and tenant assets, confirm before changing anything, and wait for long operations to finish. Includes tools that draw interactive architecture and dependency graphs.
Agent skills for administering a real UniFi network over its local API: auth, zone firewall, Wi-Fi, clients, and the site map.
Creates presentations from ideas, outlines, or existing decks, offering both editable PowerPoint (.pptx) slides and animated browser-based HTML slides. Built for Claude Code users who need slide decks.
Turns a rough product idea into a build-ready requirements document that engineers can develop from and QA can write tests against. It also finds gaps and edge cases in unfinished requirements, and generates test cases from acceptance criteria.
Multi-agent code review for Claude Code in any language or stack. It runs the repo's own typecheck, lint and tests, has several review agents read the diff in parallel, then has a separate verifier re-check each finding before it reaches the report.
Describe your goal, approve the spec, then step away, Claude and Codex loop together until it's right.
Your client wants 30 days of social content across six platforms with brand-faithful imagery, AI-generated video, and provenance signed for EU markets. You have five days.
A Claude Code hook that checks the session record before dependent steps such as file edits, notebook edits, git commits and pushes, and session stops run. It helps users verify that an agent's actions are consistent with what it recorded.
An MCP server that lets AI agents in Claude Code, Cursor, Claude Desktop and similar hosts call paid APIs for search, scraping, contact data, media generation and research from one balance, with a spend ceiling.
A Claude Code skill for managing Backblaze B2 cloud storage directly from your terminal, list files, audit buckets, clean up stale data, and review security posture. Built on the open Agent Skills specification.
A Claude skill that generates a complete handoff document of your current conversation, so another LLM (ChatGPT, a fresh Claude session, Gemini, anything) can pick up exactly where this one left off.
Real-time secret-leak guardrails for AI coding agents (Claude Code, Codex), Git hooks, and CI.
Lets you use the Kimi model from inside a Claude Code or Codex session to explain code, review changes, challenge a design, or pursue a goal. It gives an independent second opinion from a different model family.
A Claude Code plugin that structures AI-assisted development with skills, agents and commands, including standards for authoring plugin agents and commands, a commit workflow, a self-review step, and D2 diagramming.
Routes Snowflake-related prompts from Claude Code to Cortex Code CLI, covering SQL, data governance, dynamic tables and machine learning work. Includes setup skills to install the Snowflake and Cortex Code command-line tools.
Semantic code search plugin for Claude Code using hybrid vector search + BM25. 98% accuracy, 5x faster than grep.
Compound V is a Claude Code plugin for planning and building features with multiple AI models. You describe a feature, and Claude sizes it, splits it into non-overlapping pieces, and hands each to a worker in its own isolated worktree. It also includes skills for PR review and promo clips.
Stops an AI coding assistant from charging ahead on vague requests, picking the first obvious option, or making claims without checking sources. Its skills ask clarifying questions, test plans for failure, look for alternatives, verify research and review work before you rely on it.
Threat hunting command system for agentic IDEs.
A Claude Code plugin that enforces verification discipline on coding tasks, withholding completion until observable validation passes. It uses hooks and a skill with risk-proportional review and final reporting.
Edits Russian text using Maxim Ilyakhov's infostyle method. It removes filler words, clichés, bureaucratic phrasing and unsupported claims, replacing them with concrete facts and benefits, and adjusts strictness to the context.
A Claude Code plugin of agent skills for repository setup, AGENTS.md and CLAUDE.md auditing, settings and permission review, agent-compatibility checks, and designing CLIs for coding agents.
Installs and keeps AI agent assets like skills, rules, commands and agents in sync across all your projects from several Git repositories. Edit a file once and the change reaches every project through symbolic links, with no format conversion.
Shows the status of each running Claude Code session in its terminal tab, marking whether it is running, idle, or needs attention with a flashing prefix.
A coding-agent skill that applies SOLID principles and design patterns to backend code, with judgment and restraint. The goal is not to maximize the number of principles or patterns in the code.
A Claude Code skill for social media work: it builds monthly content calendars, audits a brand's social channels, and runs competitor audits. Useful for marketers and brand teams planning posts.
Extracts a Figma design system into a structured design.md file, turning design tokens into markdown that AI coding assistants such as Claude Code can use. Useful for teams that want their design tokens documented for AI-assisted development.
Claude Code skills for founders, developers, and web3 builders. This repository publishes reusable skill folders under skills//, ships stable bundle downloads through GitHub Releases, and publishes changed skills to ClawHub.
A collection of agent skills for embedded systems work, covering MCU debugging, C framework integration, charger and fuel gauge ICs, BLE GATT services, and bootloader debugging.
Automatically archives the topics of your Claude Code sessions to markdown files, so context survives session ends and compactions. You can save, list, or exclude topics, and pin preferences to carry across sessions.
A Claude Skill that runs several LLMs as a council, combining their answers through peer review and a final synthesis step.
Gives AI agents domain expertise for regulated business work, covering finance, banking, legal and sales tasks such as reconciliation, KYC risk rating and AML checks. It encodes business rules and regulatory knowledge so an agent can carry out these workflows on its own.
Gives your coding agent ready-made skills for committing code, fixing defects from a reproducible loop, improving tests, writing project docs, and driving a browser for UI checks. It works across Claude Code, Codex CLI, Copilot, Cursor and Pi.
A Claude Code hook that rewrites each user prompt so the AI first works out what is being asked before acting, keeping it aligned from the first message onward. Aimed at developers who want clearer, faster results from Claude Code.
AI coding assistant skills for building CrowdStrike Falcon Foundry apps. Build Foundry apps from a natural language prompt, API integrations, workflows, UI pages, functions, collections, AI agents, and knowledge bases, all scaffolded with the Foundry CLI.
Yoink is a Claude Code plugin that removes third-party dependencies by reimplementing only the parts a project needs. It generates and validates tests against the real library, then decomposes the dependency into internal replacements.
A library of skills for Claude Code and other AI coding agents, covering project instructions, persistent agent memory, lifecycle hooks, plugin packaging, and writing new skills.
A multi-model AI orchestration MCP server for automated code review and LLM-powered analysis.
Turn any material into a course that actually teaches. Built on learning science, not content dumps.
Gives AI coding agents design guidance for building UI, routing work across web, Apple, Android, Windows, cross-platform, and design token tasks. Covers platform idioms and design token systems.
A Claude Code skill that gives a prioritized design critique of a UI from a URL, screenshot, or component file. It covers visual hierarchy, typography, spacing, color and contrast, motion, component states, and responsiveness.
A Claude Code plugin that makes AI coding agents follow engineering discipline. Plan before coding. Verify before claiming done. Find root cause before patching. Review before merge. Skills activate based on file type and task signals, not manual toggling.
A markdown editor where your AI agent drafts and edits your writing, and you accept or reject each change in the browser before it lands. Covers blog posts, books and newsletters, with a skill for removing AI-sounding patterns from text.
A skill that gives AI coding agents workflow knowledge of the Paymob payment gateway, covering checkout, the Intention API, HMAC webhooks, SDKs, refunds, reconciliation and go-live across Egypt, the UAE, Saudi Arabia and Oman.
Gives Claude Code a supervised team of expert agents that write code, plus hooks that check quality as it is written and skills that bring in framework knowledge, so Claude follows your architecture instead of guessing.
Claude Code skills for geospatial work: inspect raster and vector files, download NAIP imagery and Overture Maps data for a bounding box, and run pre-trained AI models to detect objects in satellite imagery. Built on the geoai Python library.
Claude plans. Cursor writes. Claude reviews. A Claude Code plugin that delegates coding execution to Cursor's Composer, without ever leaving the Claude Code TUI.
Gives Claude Code tools to summarize, organize and extract insights from your SpecStory AI coding session history. It also scans for leaked secrets before commits and tracks URLs your sessions fetched.
An MCP server that gives AI coding agents persistent graph memory of decisions, context and session history, with token compression, semantic caching and session checkpointing. Works with any LLM.
Routes each prompt you type to the installed skill that fits it, with the process skill that goes with it, before Claude takes any action. You get fewer missed skills and less time spent on the skill listing every session.
Gives your Claude Code session a structured pipeline for building software, from specs and plans through code, review and pull requests. Each stage leaves a plain Markdown artifact that feeds the next, so nothing is orphaned.
Plans complex multi-step tasks, runs modules in parallel, validates each one, and retries failures while keeping memory of past results. Includes commands to check project status and validate a single module.
Project continuity for Claude Code and Codex: a roadmap beside your code, grounded handoffs, and clear task status.
A Claude Code plugin that enforces hexagonal architecture (ports & adapters) in Symfony projects. Works with both new projects (full scaffolding) and existing projects (progressive, module-by-module refactoring).
AI-powered video production suite for Claude Code. Edit, transcode, caption, analyze, generate, and create promo videos: all from your terminal.
geo-score is an open-source AI visibility (GEO) scorer for websites, by Jianrun Tech (见润科技). Formerly geo-watch: geo-watch merged into geo-score in 1.2.0 and lives on as its levels 2 and 3. Website, docs and a free online check: geo-score.dev.
Builds React and Next.js UI from design specs with Tailwind tokens, flex layout, and interactive hit areas. Fires on design-to-code syncs, paper-to-code conversion, and Tailwind token consolidation.
Builds Figma component sets and Figma variable collections (primitive and semantic design tokens) from component contract files. Use it to keep Figma designs in sync with your component and token definitions.
Gives your AI coding agent production architecture rules for FastAPI, Next.js and Go, so code it writes stays maintainable as the project grows. Adds guided debugging and checklists for new features and integrations.
Monitors Claude Code and OpenCode sessions locally in real time, showing tokens, costs, tool calls, agents, MCP activity, sessions and quotas as they run.
Helps long-form writers research a topic, score and rewrite an article to raise its viral potential, and publish the finished draft to several Chinese platforms at once. Each step is a separate tool, so you can use only the one you need.
A Claude Code plugin with specialist agents, commands and skills for software development, covering code review, architecture planning, REST API design, debugging and security patterns.
A transparent proxy that automatically summarizes older messages in Claude Code while keeping recent messages verbatim, so long sessions avoid the context limit. Requires no configuration.
Turns any text, web page or URL into a ready-to-publish HTML information card. Suited to turning articles, threads, papers and data summaries into shareable image cards for WeChat, Xiaohongshu and X.
Production-grade Claude Skills for building, debugging, and shipping n8n workflows, distilled from 100+ production workflows by an n8n Verified Creator. Give Claude Code the instincts of a senior n8n engineer.
A Claude Code plugin that aims to make Opus behave more like the Fable model. It adds skills for delegating parallel work, a gate that checks a turn before it ends, and a prose-first writing style, enforced by hooks.
Provider-neutral skills for coordinating AI agents, covering acceptance gates, context budgets for subagent handoffs, adversarial debate between options, and a Python harness for bounded, resumable agent work.
A skill for AI coding assistants that generates and modifies Metin2 client UI code.
A Claude Code workspace for product managers and designers with skills for drafting communications, sourced competitive research, epic drafts, product decisions, and an interactive learning guide. Your vault notes serve as persistent context.
A Claude plugin of game design skills for building games on any platform, covering rendering, animation, art direction and readability, combat design, and keeping AI-generated content coherent.
GIF creation skill for Claude Code with 6 generation pipelines: AI video-to-GIF (Veo 3.1), programmatic animations (Remotion), AI image sequences (Gemini/FLUX.2), animated SVG-to-transparent-GIF (Playwright), video conversion (FFmpeg), and GIF.
A Claude Code plugin with skills for product managers, covering competitive research, experiment analysis, customer feedback synthesis, decision records, and HTML slide decks for product presentations.
Generates animated, hand-drawn style architecture and flow diagrams as one self-contained HTML file, with flowing dots on connectors and a one-click export to a looping GIF.
A Claude Code skill that takes a user from raw documents (Form 16, AIS, Form 26AS, broker Tax P&L, payslips) to a fully reconciled, schedule-by-schedule computation and a portal-ready data-pack JSON, designed to be handed to another agent (or a human) that.
Tells you who is favored, the live score, and what matches are worth watching for football and basketball, using Asian-market odds. A read-only lookup that needs no account or API key.
Builds polished, interactive HTML presentation decks from a short brief, or rebuilds an existing PDF or PPTX as a single shareable file. Covers pitch, sales, launch, keynote and all-hands decks, and you can edit any element in the browser.
Quizzes you on the code Claude Code just wrote to confirm you understand it, and by default blocks the session from continuing until you pass.
Lets an engineer describe a stormwater modelling task in plain language and have the work run as a checked SWMM workflow, covering building the drainage network, assembling the model, running calibration and validation, and recording where each input came from.
Gives Claude Code skills, agents and MCP servers for building Solana applications with Helius infrastructure, covering transaction sending, NFT and asset queries, real-time streaming, webhooks, and integrations with Jupiter, DFlow, OKX and Phantom.
A Claude Code plugin that combines Superpowers workflow skills with Beads task memory, covering brainstorming, parallel agents, plan execution, and release documentation for coding agents.
Brings a local NotebookLM-style knowledge base, multi-agent deep research and other commands to Claude Code, powered by Google's Antigravity CLI (agy, Gemini models).
Persistent project memory for Claude Code that loads at session start and is organized by a schema in hot and cold layers. A memory-loop skill writes structured entries after tasks, decisions, and gotchas.
Star on GitHub if Hardstop keeps you safe! Pre-execution safety validation for AI coding agents.
A Claude Code plugin of skills and agents for AI-native development, including a codebase readiness assessment for AI agent contributors and a complexity hotspot treemap. It runs locally against your own codebase.
Gives Claude Code memory across sessions. It pulls durable lessons such as decisions, conventions and bug fixes out of each conversation and surfaces the relevant ones next time, so you stop re-explaining your project. Everything is stored locally as plain markdown.
Scaffolds an initial Claude Code setup for existing (brownfield) codebases, covering single-app and monorepo layouts, and can replace the Bash tool with structured MCP tools. Also includes a helper for headless batch scripts that run isolated sessions per item.
A control plane for running several Claude Code sessions on one machine. It shows which session is blocked on a permission prompt, which terminal is which, and which sessions are editing the same file.
Lets Claude Code hand coding tasks to isolated CLI agents such as Codex, freeze what they produce, verify it independently, and merge only changes a human approves. Includes skills for delegated implementation and debugging.
Produces content pieces end to end: drafting, fact-checking, SEO and AI-answer optimization, batch queues, and library audits to refresh stale posts. Built for content teams shipping many articles.
subscope reads Reddit for you and hands you the threads worth replying to: the people shopping for what you sell, and the questions you can answer to build authority. Run it whenever you want.
A skill that edits or reviews text to remove AI-generated writing patterns and make it sound more natural and human. Focused on Traditional Chinese as used in Taiwan.
Practical agent workflows for Python, ML, and open-source maintenance. AI-Rig turns recurring work, scoping a change, reproducing a bug, reviewing a pull request, running an experiment, or checking release readiness, into explicit workflows with specialist.
Self-hosted semantic code search for developers, combining keyword and meaning-based matching across your repositories. Adds skills that guide an AI agent to search, navigate definitions and references, and run cross-project research.
Helps you build, review, debug and cut the running cost of Atlassian Forge apps. It can scaffold a new app, check one for release readiness, diagnose errors and blank screens, and build Teamwork Graph connectors.
Scaffolds a scheduled doctor agent into your repo that audits app health on a recurring schedule and opens pull requests to fix code-fixable findings. Add new audit runbooks to an existing doctor without re-scaffolding it.
Local AI skills central library + scoped symlink enablement. One AISkills directory holds every skill source; symlink what you need into a project, the shared user-level Agents directory, or Hermes on demand.
A Claude Code skill that turns raw UX, product, and design project notes into portfolio case studies. It runs a short intake, then drafts the story in a format suited to the project and audience.
A Claude skill that turns a request, myth, or premise into a finished, narrated mythic short film in a layered cut-paper diorama style, with stacked matte cardstock worlds and a strict palette.
A Claude Code skill that hands implementation work to Sol through the Codex CLI, while Claude plans, orchestrates and reviews the diff. The model that writes the code never grades it.
Browse recent Claude Code sessions across all your projects and jump back into any of them. Shows project path, summary, git branch, model, and a ready-to-paste resume command. Works on local files only, with no network access or telemetry.
Agent skills for Julia development, including a skill for setting up and previewing Julia documentation sites with DocumenterVitepress.jl and VitePress. Works with Claude Code and other tools that support the Agent Skills standard.
Builds professional academic economics slide decks in LaTeX/Beamer. It turns a paper into a conference, seminar, or job-market talk, builds discussant slides, and polishes or shortens an existing deck.
A Claude Code plugin that catches vague prompts before they run. It detects your stack, asks targeted questions, and turns the request into a structured prompt.
An outbound prospecting pipeline of agent skills for Claude Code and Codex. Describe a customer profile in plain English and it finds matching companies, qualifies them, and returns contacts ranked by signals.
An iterative literature review workflow for research topics. It drafts search keywords, builds a seed set from OpenAlex matches, expands through citation links, and helps draft the review.
A collection of Claude Code skills for development work, covering Go with Eino and GORM, PHP with Hyperf and Laravel, plus a code simplifier for clearer code.
A skill for writing Swift code for Apple platforms such as iOS, macOS and visionOS. It covers Swift concurrency, SwiftUI, MVVM architecture, Metal GPU programming and RealityKit.
Tandem is an MCP orchestrator for Claude Code. It lets a caller start, inspect, steer, and coordinate Claude, Codex, shell, or Hermes agent sessions across multiple devices, with skills for agentic engineering workflows.
Lets Claude Code, Codex, or Cursor drive Grok Build and pull live X search results and data into the agent's context. Intended for users who already have a Grok subscription.
An agent skill that turns a topic into a designed, print-ready PDF book, with researched outline, per-chapter writing, illustrations, typesetting, and a fact-check pass.
© 2026 Flowy
Built for Claude Code · Not affiliated with Anthropic
Logos are trademarks of their owners.