hve4638/cc-toolkit
8 skills0 installsGitHub
Install
npx skills add https://github.com/hve4638/cc-toolkitSkills in this repo
1Arisaris is a set of ARIS (Autonomous Research In Sleep) skills for automating academic research. It lets an agent carry out research tasks autonomously, including unattended runs. Useful for researchers who want long-running, automated literature and investigation workflows.0installs2Corecore is the base bundle of the cc-toolkit, providing 19 general-purpose agents, hooks for rule reminders and context guarding, an MCP server with LSP and AST Grep, and common slash commands for commit/PR, review, handoff, docs, and rules. It is the foundational layer the other toolkit plugins build on. Useful as an all-in-one starting kit for agent-assisted development.0installs3Expertexpert provides advanced, heavyweight tools, including a Python REPL MCP server intended for future ralph-like workflow skills. It extends an agent with deeper execution capabilities. Useful for power users who want a Python execution backend wired into their agent.0installs4Hudhud is a heads-up display statusline that surfaces live session state, including git status, todos, context percentage, active agents, ralph, and background tasks. It keeps key signals visible while you work in the terminal. Useful for monitoring an agent session at a glance.0installs5Inlayinlay is a code-work methodology guardrail for preserving context, built around an INLAY.md file, WHY: comments, and explicit domain terms. It helps an agent retain the reasoning behind code so intent isn't lost. Useful for keeping codebases navigable and well-documented as they evolve.0installs6Oc Browseroc-browser runs an isolated headed Chromium inside Docker, invoked via oc-as <name> browser <action>, with host, VNC, and headless modes and per-project containers. It gives agents a sandboxed browser for automation and visual interaction. Useful for safe, reproducible browser automation isolated per project.0installs7Researchresearch is a set of research-workflow skills covering journal, plan, report, and commit steps. It structures how an agent investigates a topic and records what it learns. Useful for keeping research organized and producing written reports.0installs8Wikiwiki is a persistent markdown knowledge base (a Karpathy-style LLM Wiki) backed by MCP tools and SessionStart/PreCompact/Stop hooks. It captures and recalls project knowledge so context survives across sessions and compaction. Useful for retaining durable notes an agent can read and update.0installs