hamzafer/claude-code-mods

Claude Code mods: plugins built on hooks that add live lines above the prompt, guards, panes and games. Context bar, usage meter, Codex review watch, Markdown preview, Spotify now playing and more.

What it solves

Claude Code (Anthropic's terminal coding agent) keeps its inner workings mostly invisible. This project provides a set of optional plugins that make Claude Code show what it's doing: live lines above the prompt, panels with agent activity, warnings before destructive commands, and even games to pass the time. It solves the problem of not knowing how your context is filling up, what subagents are doing, when a review finishes, or whether a risky command is about to delete files.

How it works

The plugins are built on Claude Code's function hooks — hooks that fire on events like tool calls, prompt sends, and session lifecycle. Each "mod" is a small plugin that you install individually. The project includes an installer (npx claude-code-mods@latest) that shows a checklist of available mods, lets you pick which ones to enable, and installs them into your Claude Code configuration. You can also order the "bands" (the live lines above the prompt) with a simple interface. Mods are installed as plugins via Claude Code's marketplace system, and some mods need extra tools (e.g., Chrome for image rendering, gh for GitHub features).

Who it's for

Developers who use Claude Code in the terminal and want more transparency and control over their sessions. It's ideal for people who want to see context usage, monitor subagents, guard against destructive commands, or even add entertainment while waiting. The project targets users comfortable with the terminal and Claude Code, and includes community mods that are maintained by other authors.

Highlights

  • Live status lines (e.g., context-bar, token-weather, cache-clock) show context usage, token cache, and session cost above the prompt.
  • mission-control provides a live pane mapping subagents, tool calls, and the files they touch.
  • blast-radius holds risky commands like rm -r and force pushes, showing what would be deleted before proceeding.
  • switchboard routes subagents to cheaper models (e.g., from Claude Opus to Haiku) automatically.
  • Fun mods like reels (YouTube Shorts), snake, and now-playing (Spotify) add entertainment while Claude works.
  • Community mods add features like image previews, Chinese localization, and a mod store with ~2,700 mods.
  • The installer lets you pick exactly which mods to enable, with --dry-run and --pick options for automation.

Related

  • Project
    liza-mas/liza

    Liza is an accountable multi-agent system for software delivery. It pairs a behavioral contract with adversarial review, mechanical enforcement, and an engineered pipeline to turn a goal into reviewed, tested, documented code, keeping humans in charge of direction rather than execution details.

  • Project
    Varietyz/Disciplined-AI-Software-Development

    A structured methodology for building software with LLMs that replaces chat-based instructions with automated checks, formal grammars, and project-file-based constraints to prevent architectural drift.

  • Project
    mizchi/skills

    A collection of standardized, portable skills for AI coding agents, covering tasks from frontend review to cloud infrastructure, supported by an evaluation CLI for testing and iteration.

  • Project
    raysonmeng/agent-bridge

    AgentBridge is a local tool that connects Claude Code and OpenAI Codex so they can exchange messages, split tasks, and review each other's code automatically. It runs a Bun‑based daemon and a short CLI (`abg`/`agentbridge`) that launches both agents, handles turn coordination, quota hand‑off, and filters noisy output. Designed for developers who want two LLM coding assistants to collaborate without manual copy‑pasting.