Find the open-source solution that fits.

Software Tailor Skills Library packages a custom-software consulting workflow into reusable Claude skills. Its modules cover niche research, discovery questions, project blueprints, value-based pricing, proposals, feature prioritization, onboarding, demonstrations, maintenance, bug triage, and follow-on work. The ZIP packages are meant for entrepreneurs using AI coding tools to serve B2B clients, including people without a traditional development background. Revenue figures in the examples illustrate the author’s method rather than guaranteed business outcomes.

Choose from standalone skills for querying public datasets, generating charts and Sankey diagrams, or creating SVG logos and graphics. Claude Code users can install the corresponding marketplace plugins, while other Agent Skills-compatible tools can copy the desired skill directory. This repository extends an existing agent workflow rather than providing a separate data or design application.

fallow-skills gives coding agents practical command knowledge for auditing JavaScript and TypeScript projects with fallow. Natural-language requests can uncover unused files, exports and dependencies, duplication, circular dependencies, complexity hotspots, architecture violations, design-system drift, and changed-code risk. The skill also explains JSON output, requires dry runs before fixes, and uses tracing to investigate false positives. Fallow must be installed, while its optional runtime layer adds hot-path and cold-path evidence.

RuSwitcher converts the last word between any two installed keyboard layouts using a configurable key trigger, or can attempt precision-first correction as you type. It retypes the converted text directly without touching the clipboard and also works through Apple Screen Sharing. Automatic conversion and remote-desktop support are still marked as beta features.

SKILL.mk proposes a Makefile-based specification and toolset for agent skill documents. Workflows become explicit targets and dependency graphs, allowing a dedicated loader to fetch only the recipe associated with the current keyword instead of injecting the whole skill at once. Because the document is also runnable, it can be verified and audited through Git history, invocation statistics, or other checks, and a conversion script helps translate existing SKILL.md collections. This is a proposed skill-document format rather than a standalone end-user application.

Council is a native macOS app that sends one question to several cloud or local language models, streams independent answers, and lets the models critique one another anonymously. It then shows agreement, camps, outliers, and divergence so the user can decide. Users bring their own provider keys or connect Ollama and OpenAI-compatible endpoints; keys stay in the macOS Keychain, with no Council account, server, or telemetry. It requires macOS 14 or later and shows a Gatekeeper warning on first launch because the release lacks a paid Apple certificate.

MadCop is a cross-platform, local-first desktop workstation for multi-step AI work. It connects to a user-supplied OpenAI-compatible endpoint and combines persistent conversations, files, a local knowledge base, built-in tools, MCP servers, and several workflow modes in one workspace. Conversations, files, and knowledge are stored locally, while remote-model requests are sent directly to the endpoint the user configures. The documented setup is primarily source-based rather than a hosted account service.

Encrypted Notepad II keeps the familiar workflow of a basic Notepad-style editor while storing every document as a symmetrically encrypted OpenPGP ASCII-armored file. Files remain portable and independent of accounts, cloud services, proprietary databases, or network access. The full source can be built by users, while ready-made Windows, Linux, and Android binaries are sold with priority support. Encryption is implemented, but the application is still described as alpha software with unfinished UX polish.

xgrok provides a management interface for proxy access and network tunneling across HTTP(S), TCP, and UDP. It is available as a Windows or macOS desktop application, while a Docker deployment exposes a web management interface. The project also covers reverse proxying, tunnel status, startup behavior, and proxy-based access. On macOS, the application may trigger system verification or “damaged app” warnings and require the manual quarantine-removal step described by the project.

These skills let compatible agents create and inspect Weaviate collections, run keyword, vector, or hybrid searches with filters, and import data including PDFs and multi-vector records. The repository also includes end-to-end cookbooks for data explorers, chatbots, agents, and several RAG patterns, with installation through npx, Claude Code plugins, or a manual clone.

BibiGPT Skill routes video and audio tasks through a desktop CLI, remote MCP server, OpenAPI, or Codex plugin. Its workflows cover quick summaries, chapter-level deep dives, transcript extraction, article rewrites, batch processing, multi-source research, and exports to Notion, Obsidian, or local files. The remote MCP option uses OAuth without a local install, while other modes require the BibiGPT client or service endpoints.

Forge is a local-first Claude Code persona engine with separate workflows for modeling yourself and approximating another person's voice and interaction patterns. A self persona can support reflection and multi-agent decision discussions, while a person persona can be used for memory-based role dialogue. The repository includes parsers for chats, diaries, and social material, plus local persona files, validation, versioning, and decision logs. It is explicitly intended for personal memory and emotional healing, not harassment, tracking, or privacy invasion.

Quick Symlink adds common link operations to the macOS Finder context menu, avoiding repeated terminal commands. You can create links for selected files or folders, copy a source and paste its link into another directory, or move an item while replacing its original location with a symlink. After installing the app, the relevant Finder extensions must be enabled in system settings, and users can choose whether newly created symbolic links use relative or absolute paths.

HackDesk is a desktop client that wraps hackmd.io in a lightweight Tauri application, giving HackMD users a dedicated window for their existing collaborative Markdown workflow. The stable macOS build can be installed through Homebrew, but it is currently unsigned and may require manual approval on first launch. A parallel Electron beta is developing a more native workspace, search, settings, and note-management experience while the Tauri version remains the stable path.

Kaya gives Go players a browser and desktop workspace for playing games, editing variations, organizing SGF records, and reviewing positions. KataGo analysis adds win rates, move suggestions, quality graphs, and ownership heatmaps, while photo recognition can turn a physical board into an editable SGF. It runs on Windows, macOS, Linux, the web, and as an installable offline-capable PWA, making it suitable for both game review and regular study.

Awesome Claude Corporate Skills groups 166 skills across 14 business areas, including leadership, finance, HR, marketing, sales, legal, operations, engineering, product, and analytics. Teams can copy a full role or individual SKILL.md folders into Claude Code, use the repository as a Cowork workspace, or reference the instructions through the Claude API. The collection mixes custom, Anthropic-sourced, and community material, so teams can select and review what fits their standards.

Wizards of the Ghosts is an unofficial Hermes Agent skill pack that maps 123 fantasy spell and skill names to practical modes for investigation, planning, monitoring, repair, messaging, and automation. Its generated surfaces can be installed into Hermes or OpenClaw, and the same spells are exposed as slash commands in Claude Code. The pack is mainly procedural Markdown that guides reasoning and operating patterns inside an existing agent session. References to external APIs or systems do not mean the repository ships dedicated integrations for them.

Harness Starter Kit helps an agent inspect a repository and add the smallest useful set of instructions, constraints, feedback loops, memory, checks, and evaluation records. Its prompt-first commands cover doctor, adopt, review, update, and refresh, while preserving existing architecture. The kit documents operating context; it does not by itself prove better agent outcomes.