A terminal dashboard to monitor AI coding agents like Claude Code and Codex CLI in real-time.
abtop is easy to set up with strong trust signals. Check agent compatibility and use-case fit before adding it to your workflow.
gh repo view graykode/abtop --webOpen the official repository or website.
Check the README for package manager, auth, and platform requirements.
Try it in a small test task inside your agent workflow.
abtop is a tool that shows you what your AI coding assistants are doing. It displays token usage, rate limits, open ports, and other stats for multiple agents at once. It works on macOS, Linux, and Windows without needing any API keys.
abtop is a terminal-based monitoring tool inspired by btop, but designed specifically for AI coding agents. It provides a real-time dashboard that displays session information for Claude Code, Codex CLI, and OpenCode agents. You can see token usage, context window percentage, rate limits, child processes, open ports, and more. It discovers sessions from local process and file state, supporting multiple active profiles across macOS, Linux, and Windows. Features include session jumping via tmux, 12 built-in themes (including colorblind-friendly options), and a read-only design that requires no API keys or authentication. It can also detect orphan ports left by agents and warn when context windows are filling up.
Strong trust signals; still review the README and permissions before production use.
Last commit was about 9 days ago.
2545 GitHub stars indicate community interest.
17 open issues signal maintenance load.
MIT license detected.
Monitor multiple AI coding agents across different projects from a single terminal screen.
Track token usage and rate limits in real-time to avoid hitting API quotas.
Detect orphan ports left by agents that spawned servers and forgot to kill them.
View context window usage per session to know when it's filling up.
Jump directly to an agent's tmux pane by pressing Enter on its session.
abtop is read-only and does not send any data externally. However, it reads process and file state on your machine, which could expose sensitive information if the terminal output is shared.
2,545
Stars
241
Forks
17
Issues
MIT
License
A full-stack AI Red Teaming platform for securing AI ecosystems with comprehensive scanning and evaluation.
AI agent security scanner that detects vulnerabilities in agent configurations, MCP servers, and tool permissions.
A bridge between Streamable HTTP and stdio MCP transports, enabling flexible MCP server connectivity.
1 security/trust notes recorded.
Setup difficulty is 1/5.