◈AI Codex
Claude CodeCore Definition

Monitor tool: real-time background process streaming in Claude Code

In brief

The Monitor tool spawns a background process and streams its stdout line-by-line into the conversation without blocking Claude Code's main thread.

3 min read·Monitor Tool

Contents

♡Sign in to save

The Monitor tool is a Claude Code capability that spawns a background process and streams its stdout output into the conversation in real time, without blocking the main thread. Introduced in Claude Code v2.1.98 (April 9, 2026), it enables patterns like watching logs while writing code, or tailing a test runner while implementing the feature it is testing.

What it does

When you use the Monitor tool, Claude starts a subprocess — a running server, a build process, a log tail — and receives its output as a stream of notifications. Claude can act on that output: reading error messages as they appear, noticing when a server starts successfully, catching a test failure mid-run.

This is qualitatively different from running a command and waiting for it to complete. Monitor gives Claude visibility into a running process without that process blocking everything else.

Practical patterns

Watch a dev server while you code. Start your dev server with Monitor and ask Claude to fix a bug. As you make changes, Claude can see the server output — compilation errors, runtime warnings, request logs — without having to stop what it is doing and run a separate check.

Tail logs while debugging. If you are chasing a production issue and need to watch logs while making changes, Monitor lets Claude observe both simultaneously.

Gate on a process starting. "Start the database, then run the migrations when it is ready." Monitor lets Claude wait for the "ready" signal in the process output rather than guessing a sleep duration.

The design

Monitor is non-blocking because it uses a background subprocess architecture. Each line of stdout from the monitored process arrives as a notification — Claude reads it when it reads it, but the main conversation thread does not stall waiting for the process to finish or produce output.

For long-running processes (like a dev server you keep running for the duration of a session), this is the right model. The process runs; Claude monitors it; both continue independently.

When not to use Monitor

For commands you just need to run and finish — a test suite, a build, a script — the standard Bash tool is simpler. Monitor earns its value when you need ongoing visibility into a running process, not a one-shot command execution.

Further reading

Related tools

Claude vs GPT-4

Side-by-side comparison across code quality, context, debugging, and cost.

See comparison →

Claude API Cost Calculator

Estimate your monthly spend by model, message volume, and caching strategy.

Calculate your cost →

Weekly brief

For people actually using Claude at work.

What practitioners are building, the mistakes worth avoiding, and the workflows that actually stick. No tutorials. No hype.

No spam. Unsubscribe anytime.

What to read next

All articles →
How It Works·7 min

Claude Code routines: automate workflows on a schedule

A routine is a Claude Code automation that runs on a schedule, from an API call, or in response to a GitHub event — without your laptop open. Here's what they are and where they fit best.

Failure Modes·6 min

Claude Code anti-patterns: what to stop doing

The failure-mode companion to every Claude Code setup guide. CLAUDE.md that's too big, MCP servers left running, sessions you never close, skipping Plan Mode, asking quick questions in the main thread — here's what each one costs you and how to fix it.

How It Works·6 min

Parallel agents in Claude Code: the desktop redesign

Claude Code's desktop app was rebuilt for running multiple coding tasks at once. A new sidebar manages sessions across repos, an integrated terminal and diff viewer replace external tools, and side chat lets you branch conversations without interrupting ongoing work.