How to Monitor Claude Code Agent Status and Progress with Claude HUD

Mar 21, 2026

Problem

I was working on a large codebase refactor with Claude Code. I had asked Claude to explore the authentication module while also planning a database migration. After a few minutes, I realized I had no idea what was happening. Were both tasks still running? Which model was being used? How long had they been working?

Claude Code can spawn multiple subagents that work in parallel. An explore agent searches the codebase. A plan agent designs implementations. A tdd-guide agent writes tests. But these background tasks are invisible. I couldn’t see what was running or track progress.

What I Tried First

I looked at the Claude Code interface. It showed the current model and context usage, but nothing about running agents. I tried scrolling through the transcript, but that’s messy when you have multiple parallel operations. I needed something that showed agent status at a glance.

Then I found Claude HUD. It has an agent tracking feature that shows exactly what I needed.

The Solution

Claude HUD displays running agent status in a dedicated line. Here’s what it shows:

◐ explore [haiku]: Finding auth code (2m 15s)

This tells me:

Agent type: explore - fast codebase exploration agent
Model: haiku - using the Haiku model for speed
Task: Finding auth code - what the agent is doing
Elapsed time: 2m 15s - how long it’s been running

Enabling Agent Tracking

Agent tracking is disabled by default. I enabled it through the configuration command:

/claude-hud:configure

This opens a guided setup. I selected “Agents & Todos” to enable both features. Alternatively, you can edit the config file directly:

{
  "display": {
    "showAgents": true
  }
}

After enabling, the agents line appeared whenever Claude spawned a subagent.

How It Works

Claude HUD parses the transcript file for Task tool calls. Each Task call represents a subagent invocation. The HUD extracts:

Agent type from the subagent_type field
Model from the prompt content
Description from what the agent is doing
Elapsed time by comparing start time to current time

The display updates every ~300ms, so you see progress in near real-time.

Agent Types I’ve Seen

explore          - Fast codebase exploration
plan             - Implementation planning
general-purpose  - General coding tasks
tdd-guide        - Test-driven development
code-reviewer    - Code review after writing
security-reviewer - Security analysis

Custom agents defined in skills also appear with their configured names.

Multiple Agents Running

When multiple agents run simultaneously, they appear on the same line separated by pipes:

◐ explore [haiku]: Searching tests | ◐ plan [sonnet]: Designing API

This happened when I asked Claude to research a pattern while also planning implementation. I could see both agents working in parallel, each with its model and task.

The elapsed time helped me understand when an agent was taking too long. If an explore agent ran for 5+ minutes, I knew something was stuck.

Why This Matters

Without agent tracking, parallel workflows are opaque. You don’t know:

Which agents are still running
Whether a task finished or got stuck
How long background work has been happening
What model each agent is using

With tracking, I can:

Monitor progress without interrupting
Decide if an agent needs cancellation
Understand model selection for cost awareness
See the full picture of concurrent operations

Real Example

I asked Claude to refactor a module and write tests simultaneously. Here’s what I saw:

[Opus] │ my-project git:(main*)
Context █████░░░░░ 45%
◐ plan [sonnet]: Designing refactored module (1m 30s)
▸ Refactor auth module (1/3 tasks)

The HUD showed:

Plan agent designing the refactor (Sonnet model, running 1m 30s)
Todo progress showing 1 of 3 tasks complete

I waited. After 2 minutes, the plan completed and implementation started. The agent line disappeared, replaced by tool activity showing edits happening.

Configuration Options

Agent tracking works with other HUD features:

Option	Config Key	Default
Show agents	`display.showAgents`	`false`
Show tools	`display.showTools`	`false`
Show todos	`display.showTodos`	`false`

I enable all three for full visibility into Claude’s activity. The combination shows what tools are being used, what agents are running, and how tasks progress.

Summary

Claude HUD’s agent tracking provides visibility into parallel subagent workflows. You see which agents are running, their model, task description, and elapsed time. This makes complex multi-agent operations transparent instead of invisible.

To enable it:

Run /claude-hud:configure
Select “Agents & Todos”
Restart Claude Code

The agents line appears automatically when subagents are spawned. No extra commands needed during normal use.

Final Words + More Resources

My intention with this article was to help others share my knowledge and experience. If you want to contact me, you can contact by email: Email me

Here are also the most important links from this article along with some further resources that will help you in this scope:

👨‍💻 Claude HUD GitHub

Oh, and if you found these resources useful, don’t forget to support me by starring the repo on GitHub!