Start with the job, then find an open-source answer

Palkan's skills help coding agents examine Rails applications through the layered architecture principles described in Layered Design for Ruby on Rails Applications. The workflows can audit service objects, model callbacks, god objects, layer boundaries, and code changes, then support incremental refactoring plans and specification tests. Claude Code users get packaged commands and sub-agents through the plugin, while the underlying SKILL.md content and references can also be used from environments such as Codex.

Superpowers Lab hosts experimental Claude Code skills for tasks that are awkward to handle manually. Current examples include finding semantically duplicated functions, invoking MCP tools on demand without permanent configuration, and controlling interactive programs through tmux; the skills are still being refined and may change.

Paperling is a desktop Markdown reader and editor that opens ordinary `.md` files without requiring a vault or plugin system. It provides source, reader, and split views with live rendering for math, chemistry, and Mermaid diagrams. An optional bring-your-own-model assistant can answer questions or propose inline edits that users accept or reject. Releases are available for Windows, Linux, and Apple Silicon macOS, but current builds are unsigned and may trigger operating-system warnings on first launch.

paper2patent provides prompt templates and an agent skill for transforming a research paper into a Chinese invention-patent application draft. It guides the model through the abstract, up to ten claims, the specification, and black-and-white flowcharts or block diagrams under explicit formatting rules. Supporting scripts can generate figures, export DOCX or PDF files, and check whether a draft introduces content not supported by the source paper. The workflow can be loaded into Codex, Claude Code, Cursor, Windsurf, and similar skill-aware tools.

Install awesome-novel in Claude Code, start it from a novel folder, and let a coordinator guide worldbuilding, character setup, volume and chapter planning, drafting, cleanup, review, and archiving. The project keeps settings, progress, and finished chapters in editable Markdown files, so authors can inspect or revise them directly. It supports step-by-step approval or a hands-off mode; personal use is free, while commercial use requires separate permission.

Swift Testing Pro is an agent skill that helps Claude Code, Codex, Cursor, Gemini, and similar assistants write and review tests using Swift Testing. It covers @Test, #expect, #require, parameterized tests, traits, exit tests, confirmations, and other areas where language models often produce incomplete or outdated code. Once installed, it can be triggered by command or natural language and focused on a specific goal, such as parameterizing a suite or adding exit tests. It is intended to improve testing work inside an existing Swift project.

CoWork OS is a GUI-first desktop application with a companion `cowork` CLI for running AI-assisted work in one local workspace. It brings coding, email, research, browser tasks, documents, spreadsheets, presentations, agent sessions, and automations into a shared operator console instead of separate apps. Long-running tasks can be organized through session timelines, boards, teams, plugins, and multiple model providers, while normal local CLI use shares the desktop configuration. It is aimed at users who want a broad, locally governed environment for complex AI workflows.

qiaomu-design is a Claude Code skill that begins with a product and interface diagnosis, then creates four deliberately different browser-viewable directions for comparison. After a user selects one, it implements the design using either the existing product system or a newly defined combination of typography, color, layout, motion, and density. It requires Claude Code, and the project recommends Claude or GLM 5.2 when visual quality is the priority.

mcp-google-map exposes Google Maps functions to AI clients through MCP or a direct CLI, covering geocoding, place search, directions, distance matrices, elevation, time zones, weather, air quality, and static maps. It can run as a self-hosted stdio server, and you may register only selected tools to reduce context use. A Google Maps API key is required.

NTM turns tmux into a local control plane for multi-agent development. It can spawn and label sessions, route work across panes, coordinate assignments and file reservations, protect risky operations with approvals, and preserve checkpoints and audit trails; robot JSON and local APIs make the same controls usable from scripts.

arXiv Review Paper Harness is a portable Agent Skill for producing machine-learning and AI review papers with Codex or Claude Code. Starting from a topic, it organizes literature discovery, a human approval gate, issue-based drafting, citation-by-citation verification, prose refinement, and LaTeX compilation into an IEEEtran project with a PDF and BibTeX file. It requires Python, a working LaTeX toolchain, and web access for citation checks, and it does not begin writing the paper body until the plan is approved.

Use a desktop interface on macOS or Windows to discover, install, update, remove, and synchronize skills across coding agents. Its scanner checks for destructive commands, remote execution, credential exposure, Unicode tricks, disguised file types, and multi-step or cross-skill attack chains, with hard blocking for selected high-risk findings. The app also discovers and manages CLI tools installed through common package managers.

Equip coding agents with 37 plugins covering SAP BTP, CAP, Fiori and UI5, ABAP, HANA, Analytics Cloud, Datasphere, AI Core, and related workflows. Relevant guidance activates from project context. The skill content is portable, but Claude-specific agents, hooks, commands, and MCP configuration act only as guidance or manual validators in clients that do not support those features natively.

Agent Sandbox Skill gives Claude Code, Codex, and Gemini CLI managed E2B environments for engineering work that should not touch the local filesystem or production systems. Agents can install dependencies, run untrusted code, scaffold and build full-stack applications, expose a hosted port, and perform Playwright-based visual checks. Commands cover small sandbox tasks as well as plan, build, host, and test workflows, with lifecycle tools that preserve a sandbox across turns. An E2B API key is required, and compute and hosted previews depend on the E2B service.

hcom links coding agents running in separate terminals so they can send messages, report status, and spawn one another. It supports several popular CLIs through a single Rust binary with no background service, letting you split work, collect results, or ask one agent to check another’s changes.

Firebase Agent Skills is a collection of packaged instructions and scripts that help AI coding agents understand and work with Firebase. The bundle follows the open Agent Skills format and can be installed through the Agent Skills CLI, as a Gemini CLI extension, or through Claude and Codex plugin marketplaces. Manual copying is also documented for other compatible tools. It is not a standalone Firebase management application; its value depends on an agent environment that can load and apply the skills.

Plan a deck and generate Reveal.js HTML with tailored themes, multi-column layouts, callouts, charts, speaker notes, animations, and custom CSS. The workflow checks for slide overflow and reviews screenshots, while a browser editor lets you revise text without touching the markup. Presentations open without a build step; overflow checks and PDF export require the included Node dependencies.

yupi-skill packages the programmer Yupi's decision patterns, communication style, and selected experience into an agent skill. Once installed, an agent can answer questions about learning paths, job offers and interviews, AI-assisted coding, resume revision, technology choices, and entrepreneurship in an approximation of that perspective and tone. Identity and decision rules, voice patterns, and knowledge sources are stored separately and loaded progressively. Installation instructions are provided for Claude Code, Cursor, OpenClaw, and compatible environments.