Find the open-source solution that fits.

Tactus is a browser extension for Chrome, with a temporary Firefox debug-install path, that combines AI chat, page extraction, Agent Skills, and HTTP MCP tools in a sidebar. Users can import skills with instructions, scripts, and resources so the agent can understand the current page, call external tools, and automate repeatable workflows; first-time script execution requires confirmation. Conversation history and skill files are stored locally, but model and MCP calls still connect to user-configured external services through endpoints, credentials, or OAuth.

LinkedIn Skills helps agents handle the content workflow from drafting posts, comments, and replies to auditing copy, reducing common AI patterns, improving profiles, and planning a posting cadence. Core writing runs inside supported agents such as Claude Code or Codex. Reading LinkedIn data can optionally use Apify, while automated publishing requires a separately configured Publora account and API credentials.

SwiftData Pro is an agent skill for Claude Code, Codex, Cursor, Gemini, and other coding assistants. It guides work with @Model, @Query, predicates, indexes, migrations, relationships, iCloud synchronization, and newer API behavior, focusing on errors and outdated patterns that language models commonly miss. After installation, users can invoke it to review current API usage, identify useful indexes, improve model design, or help enable cloud support. The skill works inside an existing Apple-platform project rather than as a standalone database tool.

VueUse Skills is an experimental Agent Skills package that helps coding assistants select and apply VueUse composition utilities in Vue or Nuxt projects. It uses progressive disclosure, loading function overviews first and detailed usage or type declarations only when needed, which reduces context use and invented APIs while remaining offline-first. The target project still needs VueUse installed, and teams can override invocation policies in prompts or AGENTS.md.

MathLens is a Cursor Agent Skill that accepts a math problem as text or an image and carries it through analysis, visual explanation, storyboarding, voiceover generation, and Manim animation. Its workflow validates audio timing and writes synchronization details back into the storyboard before rendering. The repository includes initialization, TTS, structure-checking, validation, and rendering scripts for producing step-by-step teaching videos.

These skills teach coding assistants to use the Registry Broker API for live agent discovery across multiple protocols, identifier resolution, chat, and registration of agents or skills. Basic search and statistics are available from the CLI, while chat, registration, and some publishing flows require the relevant API credentials, ownership verification, or account balance.

Coro Code is a Rust terminal coding agent with an interactive UI plus built-in shell and file tools. It runs on macOS, Linux, and Windows and currently centers on OpenAI or OpenAI-compatible endpoints. Native Anthropic and Google support, tool-call permissions, and sandboxing are still listed on the roadmap, so users must set their own execution boundaries for now.

Repo-Docs gives coding agents a repeatable way to maintain living project documentation beside the source. It creates an orientation page, a walkthrough of one real behavior, a responsibility-focused code map, concept and evidence pages, a glossary, and change records with sync anchors. The output is designed to explain how the repository actually works, not merely list files, dump APIs, or preserve chat transcripts.

Tonkatsu Box is a cross-platform personal collection manager for games, movies, television, anime, manga, visual novels, books, and comics. Users can search external catalogs, import existing libraries, track status and ratings, build visual boards or tier lists, and continue using the app offline after setup. Local-network device sync copies the entire database and replaces the receiving device instead of merging changes, while macOS and Linux support is currently described as experimental.

This skill collection guides Nature-oriented manuscripts from initial structure through revision, citation checks, figure planning, statistical reporting, submission review, and rebuttal responses. It emphasizes claim-to-evidence alignment rather than surface-level polishing and can be installed for Codex or Claude Code. The workflow targets selected scientific paper types and does not replace a journal's official author instructions.

Notepanda is a deliberately simple cross-platform text editor with basic editing, line numbers, syntax highlighting, and a preview panel for Markdown and HTML. Files can be opened by drag and drop or passed directly when launching the app from a shell. Releases are available for Windows, macOS, and Linux, but the author presents it as a small learning project rather than a full replacement for mature professional editors.

Mercury Skills collects 132 SKILL.md playbooks across 23 categories in a searchable registry with detail pages, trending rankings, and local bookmarks. Skills can be installed through the Mercury CLI with version pinning and rollback safeguards, or copied directly into an agent’s skill directory. The same format is intended for Mercury, Claude Code, Codex, OpenClaw, Hermes, Cursor, and Gemini CLI, while public JSON feeds let other tools consume the catalog.

Claude AI Music Skills organizes album concepts, sourced research, lyric writing, Suno prompts, stem processing, mastering, and release preparation through 53 skills and an MCP tool server that tracks project state. Music generation still happens in Suno, and the plugin requires macOS or Linux with Python 3.11 or newer; multi-track sessions can consume Claude usage limits quickly.

HeyGen Skills chains three agent workflows: creating a reusable digital twin from a photo, turning ideas or source material into scripted avatar videos, and translating finished videos with dubbed voice and lip synchronization. The skills can use the HeyGen CLI with an API key or a remote MCP connection through OAuth. CLI usage is billed as direct API consumption, while MCP uses credits from an existing HeyGen plan. Both paths require a HeyGen account.

This benchmarking toolkit runs ARC-AGI tasks through configurable adapters for providers such as OpenAI, Anthropic, and Gemini, with rate limiting, retries, saved submissions, and scoring built in. It supports both single-task debugging and asynchronous batch runs. Full evaluations require separately cloned ARC-AGI data, and hosted models require their respective API credentials.

Velox brings a Tauri-style desktop application model to Swift, connecting HTML or modern web frontends to Swift backends through typed IPC. Its CLI can initialize projects, run watched development sessions with optional frontend HMR, build releases, and create macOS application bundles; simple static interfaces do not require Node.js. The implementation still compiles a Rust FFI layer around Tao and Wry, and the newer Swift-first runtime remains incomplete, so the event-loop APIs are currently the primary path.

Claude Code Trace turns session files from the local Claude projects directory into readable conversations. It supports searching by user message, expanding tool and MCP calls, inspecting timestamps and token counts, and following active sessions in real time. The viewer is available as a desktop app, web interface, or terminal UI and does not require sending traces to an external service.

Lovcode gives developers a desktop workbench for launching Claude Code, Codex, terminal, and other AI coding sessions from a selected project. It indexes local histories and exposes tool calls, linked files, generated images, and usage details, making previous work easier to find and resume. Users can copy handoff context between agents and manage commands, MCP servers, skills, hooks, runtimes, and model providers from one interface. Desktop releases target macOS, Windows, and Linux.