Claude HUD
A Claude Code plugin that puts context usage, active tools, running subagents and todo progress into the statusline
Low risk
We rate an entry low when it mostly gives the agent instructions and reference material.
Why this level
- The plugin only reads session data and draws the statusline; it runs no commands and writes to no external systems
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/claude-hudDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Run one by one in the Claude Code chat
/plugin marketplace add jarrodwatts/claude-hud
/plugin install claude-hud@claude-hudInstall the plugin: /plugin marketplace add jarrodwatts/claude-hud, then /plugin install claude-hud and /reload-plugins. Then run /claude-hud:setup to wire up the statusline and /claude-hud:configure to adjust the lines.
Other ways from the author
/plugin marketplace add jarrodwatts/claude-hud
/plugin install claude-hud
/reload-plugins
/claude-hud:setupAfter install /claude-hud:setup wires the statusline, /claude-hud:configure adjusts the lines.
This is third-party code. Review the repository files before installing.
What it does
The plugin keeps a live session summary in the Claude Code statusline. By default it shows two lines: the model, a provider label when it is confidently identified, the project path and the git branch, then a bar for context usage and rate limits. Token figures come straight from Claude Code rather than being estimated, and the bar scales to the context window size, up to million-token sessions. Optional lines add tool activity, subagent status and todo progress, all read from the session transcript. It runs through the built-in statusLine API, with no separate window and no tmux, and re-renders after each interaction.
Who it is for. For developers who want to see the state of a Claude Code session right in the terminal.
Good fit when
- You want to watch context usage and compact history in time
- You want to see which tools and subagents are running right now
- You want todo progress and rate limits without a separate window
Not a fit when
- You work outside Claude Code, in an agent without the statusLine API
- Your statusline already runs a script you do not want to replace
Example request
Change the claude hud statusline layout and add the todo progress lineLimitations
The plugin targets Claude Code and its statusLine API, it displays what the agent reports about the session and adds no new capabilities to the agent itself. Windows setup needs Node.js LTS. On older Claude Code versions the statusline appears only after a restart, and on Linux with a separate filesystem for tmp the plugin install could previously fail with an EXDEV error.
How to disable. Remove the plugin via /plugin, then delete or replace the statusLine entry in the Claude Code settings file that /claude-hud:setup added.
Security check
- The plugin only reads session data and draws the statusline; it runs no commands and writes to no external systems
README in short
The README describes the Claude HUD plugin and a three-step install: add the marketplace, install the plugin, run /claude-hud:setup to configure the statusline. By default it shows two lines with the model, project path, branch and context and usage bars, and /claude-hud:configure enables tool, subagent and todo lines. It also covers terminal install via the Claude Code CLI and fixes for common Windows and Linux issues. A Chinese README translation is included. MIT licensed.
FAQ
Are the context numbers accurate?
The plugin uses token data straight from Claude Code rather than estimating it, and scales the bar to the context window size, including million-token sessions.
Do I need a separate window or tmux?
No. The plugin uses the built-in statusLine API and prints the line in the terminal itself.
Related
A skills library that gives coding agents a development process: brainstorming, planning, TDD, subagents and code review
Skills for real engineers by Matt Pocock
Skills For Real Engineers
Small composable skills for engineering with agents: plan grilling, TDD, bug diagnosis, code review and architecture
GitHub toolkit for spec-driven development: the specify CLI adds agent commands and skills to a project, from principles to implementation
Reference MCP servers
Model Context Protocol servers
Official reference MCP servers: Filesystem, Fetch, Git, Memory, Sequential Thinking, Time and Everything