Claude Code Agent Monitor 🤖 Agent Open source
Self-hosted real-time dashboard for Claude Code and Codex sessions, agents, tools and cost
- GitHub stars
- 1.0k
- Stars this week
- +23
- Forks
- 244
- Licence
- MIT
- Last push
- 2026-10-02
- Maintainer
- hoangsonww
git clone https://github.com/hoangsonww/Claude-Code-Agent-Monitor.gitThird-party subagents & agents run with your permissions. Read the source before installing, and prefer pinned versions.
Works with
About Claude Code Agent Monitor
What it does
Claude Code Agent Monitor is a self-hosted dashboard for watching Claude Code and Codex sessions as they happen. It installs hook handlers that forward session, tool-use, stop and subagent events to a local Express server, which stores them in SQLite and pushes updates to a React interface over WebSockets. It also reads existing transcripts from ~/.claude and ~/.codex/sessions, so past sessions can be imported and subagent tool calls that emit no hooks are still attributed. The result is one place to see which agents are working, waiting or stuck, what tools they called and roughly what each session cost.
What is inside
- Live views: an overview dashboard, a Kanban board of agents and sessions, an activity feed and per-session detail with subagent trees
- Analytics: token use, tool frequency, activity heatmaps, cost estimates from editable pricing rules, and workflow diagrams
- Operations: rule-based alerts, webhooks, browser push notifications and a statusline script
- A Claude Config Explorer that lists installed skills, subagents, commands and plugins
- A local MCP server that lets an agent query the dashboard, read-only by default
- Optional Electron desktop app, Docker Compose stack with Prometheus and Grafana, and remote collection over SSH
Works with
Claude Code is the primary target and Codex support is marked beta. The README notes that Cursor sessions stored under the same paths are counted too, without being told apart.
How to install
git clone https://github.com/hoangsonww/Claude-Code-Agent-Monitor.git
cd Claude-Code-Agent-Monitor
npm run setup
npm run install-hooks
npm run dev
Maintenance and safety
The project is MIT-licensed and updated very frequently. The hook installer edits ~/.claude/settings.json and ~/.codex/hooks.json. The dashboard can also launch claude and Codex runs itself, and it stores transcript content locally, so keep it bound to loopback. The MCP server needs explicit flags before it can change or delete data. It needs Node.js 22 or newer.
Who should use it
Developers who run many or long agent sessions, or use subagents heavily, and want visibility into activity, tokens and cost without sending data to a hosted service.
Pros
- Live view of sessions, subagent trees, tool use and estimated cost
- Imports existing Claude Code and Codex transcripts
- Local-first with SQLite; MCP server read-only by default
Cons
- Large, fast-moving codebase that needs Node.js 22+
- Hook installer edits your global Claude and Codex settings
Similar subagents & agents
All agent workflows & frameworks →Spec Kit 🤖 AgentFree
GitHub's toolkit for spec-driven development with AI coding agents
Task Master 🤖 AgentFree
AI task management that turns a PRD into tasks your coding agent works through
Fast Agent 🤖 AgentOpen source
Python framework for building, orchestrating and evaluating MCP-native AI agents
OpenSpec 🤖 AgentFree
Lightweight spec-driven development: agree on changes before the agent codes
CC Safety Net 🤖 AgentOpen source
Pre-execution guard that blocks destructive commands and secret access for coding agents
BMAD Method 🤖 AgentFree
Agile AI-driven development with analyst, PM, architect, developer and UX agents