Collection or directoryBackend developmentClaude Code
agent-skills
@supabase·TypeScript
Give coding agents Supabase guidance for development, security, and Postgres tuning.
Project overview
Equip AI coding agents with Supabase guidance for databases, Auth, Edge Functions, Realtime, Storage, client libraries, and server-rendered integrations. Install the full collection or individual skills, then use them for session and RLS troubleshooting, schema reviews, migrations, query tuning, and index decisions across many compatible agents.
Run resumable, phase-guarded coding workflows and evaluate reusable agent skills.
Project overview
Comet organizes coding work into guarded phases from specification through build, verification, and archive, recording state so an interrupted task can resume without reconstructing its history. It also helps create, compose, distribute, and evaluate reusable skills across many coding agents. The CLI requires Node.js 20 or newer.
Schedule and steer many terminal agents across your own machines.
Project overview
AgentsMesh provides a control layer for running terminal-based coding agents across a fleet of self-hosted runners. Each agent receives an isolated Git worktree, credentials, branch, and live terminal stream, while a shared console adds scheduling, autopilot, collaboration channels, and ticket tracking. Its BSL 1.1 license requires a commercial license for production use until the stated change date.
Turn completed AI-assisted coding work into short deliberate learning exercises.
Project overview
Learning Opportunities is a Claude and Codex skill that turns significant coding work into optional 10–15 minute learning exercises. After architectural changes, new modules, refactors, or unfamiliar patterns, it uses prediction, generation, retrieval practice, and spacing to build lessons from the user's own project. The agent pauses for the developer's answers instead of completing the exercise itself, and an optional companion orientation skill can prepare lessons for an unfamiliar repository.
Use a Codex-style terminal agent with multiple cloud and local model providers.
Project overview
Open Codex CLI forks the original Codex CLI while adding support for OpenAI, Gemini, OpenRouter, Ollama, and compatible endpoints. It uses the Chat Completions API so users can choose a broader range of models, then runs through a familiar terminal workflow after provider credentials are configured. The project is experimental and may change incompatibly.
Collection or directoryResearch and knowledgeClaude Code
agent-skills
@apify·Python
Scrape websites, build Apify Actors, and deliver structured data with coding agents.
Project overview
Equip Claude Code, Cursor, Codex, Gemini CLI, and other agents to choose Apify Actors for scraping social, search, maps, commerce, and review platforms. Separate skills help create, debug, and deploy new JavaScript, TypeScript, or Python Actors, wrap existing scripts as Actors, generate output schemas, and integrate runs through the SDK or REST API. The collection supports both one-off data requests and maintainable automation workflows.
Let a standalone coding agent research, plan, and execute multi-step development work.
Project overview
RA.Aid moves a development task through dedicated research, planning, and implementation stages. It can inspect a codebase, run shell commands, edit files, ask for human input, and optionally delegate edits to a separately installed aider. Because it can execute commands and change code autonomously, the project recommends using version control and reviewing the Git diff before committing.
Collection or directoryMobile app developmentAndroid
skills
@expo·Shell
Guide AI agents through building and debugging Expo apps
Project overview
Expo Skills gives compatible coding agents focused guidance for building, deploying, upgrading, and debugging Expo apps. The collection explains common Expo and EAS workflows plus React Native, iOS, and Android constraints, and can be installed through official plugin marketplaces or the skills CLI for different agents.
Study how to build document indexing, retrieval, and cited memory for AI applications.
Project overview
Kernel Memory is a reference implementation for turning documents into an AI memory layer through extraction, chunking, indexing, retrieval-augmented generation, and source citations. It can run as an asynchronous service or as an embedded component in .NET applications. The repository is archived and explicitly presented as unsupported research code, so it should be used for learning and architectural reference rather than production deployment.
Collection or directoryFrontend developmentClaude Code
Skills
@MengTo·Python
Reuse interface-design and web-building workflows across Codex, Claude, Cursor, and other AI agents.
Project overview
This collection offers 75 skills for designers and builders working on web interfaces, animation, asset sourcing, and prompt creation from videos or existing pages. Its plain Markdown, folder-based format lets compatible agents load only the relevant SKILL.md workflow instead of depending on a single vendor-specific runtime.
Research literature and draft cited papers, reviews, reports, posters, and grant proposals.
Project overview
Claude Scientific Writer combines literature search, citation management, peer review, and specialized scientific-writing skills for papers, reviews, reports, posters, slides, and grant proposals. It can run as a Claude Code plugin or through its Python package and API. Research lookup in version 2.13 and later requires a PARALLEL_API_KEY, while some image-generation features optionally use OpenRouter.
MCP serverResearch and knowledgeModel Context Protocol
DevDocs
@cyberagiinc·TypeScript
Crawl technical documentation into a local knowledge base that AI tools can query.
Project overview
DevDocs provides a graphical workflow for crawling technical documentation sites, selecting depth and pages, cleaning the content, and organizing it as Markdown or JSON for querying through an included MCP server. It supports link discovery, parallel crawling, caching, and local self-hosting, with Docker as the recommended setup for the full service stack. The public repository is explicitly no longer publicly maintained, and its Windows setup is described as experimental, so adopters should account for maintenance risk.
Aggregate coding-agent sessions and delegate work across multiple agents.
Project overview
Codeg brings sessions from Claude Code, Codex, OpenCode, Gemini CLI, and other coding agents into one workspace. A primary agent can delegate parts of a task to independently running sub-agents, while built-in Git worktree flows support parallel development. The workspace also includes project scaffolding and live Office document editing. It can run as a desktop application, a standalone server, or a Docker deployment, but the underlying agents and their local credentials or configuration are still required.
Turn repeated workflows into evaluated, portable agent skills.
Project overview
Yao Meta Skill converts workflows, prompts, transcripts, notes, or runbooks into reusable skill packages with explicit triggers and artifacts. A platform-neutral skill model can be compiled for OpenAI, Claude, generic Agent Skills, and VS Code-oriented targets, then checked through behavioral evaluations, review gates, and release evidence.
Solve software defects through structured localization, repair, and patch validation.
Project overview
Agentless replaces a continuously interactive agent with a three-stage process: fault localization, candidate repair, and patch validation. It narrows an issue from files to classes or functions and then exact edit locations, samples several diff patches, and reranks them using regression and reproduction tests. The project mainly targets SWE-bench-style repair experiments and requires a Python 3.11 environment plus an OpenAI API key.
Give each AI agent an always-on computer, memory, and chat channels.
Project overview
Memoh gives each agent an isolated workspace with its own filesystem, desktop, browser, network, tools, and long-term memory. You can run its built-in agent with your own API keys or host existing Claude Code and Codex agents, then reach them through Telegram, Discord, Lark, WeChat, the web interface, and other channels. Agents can keep running scheduled and browser-based work on a server after your laptop is closed. The full platform is currently self-hosted; the zero-setup Memoh Cloud offering is still on a waitlist.
Use Codex Autoresearch when a repository task has a numeric outcome, such as fewer test failures, better coverage, lower latency, or a smaller binary. The skill confirms the target, editable scope, metric command, regression guard, and run mode, then lets Codex try one focused change per iteration. Improvements are committed and retained; failed or regressive trials are reverted through Git, with an auditable event history. It can run interactively or in the background, but each experiment requires a clean named Git branch and sufficient access to create and revert commits.
Structure external reconnaissance for authorized red-team and bug-bounty engagements.
Project overview
Claude-OSINT pairs a strategic methodology skill with a tactical reconnaissance arsenal for authorized security work. It helps an agent build an asset graph, investigate web, cloud, identity, and exposed-secret surfaces, and organize findings by scope, confidence, and severity. The repository also includes read-only credential validators, secret-scanning helpers, engagement examples, and reporting templates for red teams, attack-surface reviews, and bug-bounty workflows.