Subagent Dashboard for Openclaw

A centralized web dashboard for real-time monitoring, transcript viewing, and management of active OpenClaw subagents.

austindixson
v1.0.0
Mar 8, 2026
0
596
2

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install subagent-dashboard

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install subagent-dashboard using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Subagent Dashboard?

The Subagent Dashboard is a vital component of the Openclaw Skills ecosystem, providing a graphical interface to track the lifecycle of your AI agents. It eliminates the need to manually tail log files by aggregating data from the subagent-tracker into a clean, responsive Flask-based web UI. Developers can use this tool to gain deep visibility into how the main orchestrator and its delegated subagents are performing at any given moment, ensuring that complex autonomous workflows remain on track.

Subagent Dashboard Use Cases

  • Monitoring long-running autonomous tasks across multiple subagents to ensure system health.
  • Debugging complex agent behavior by viewing live activity transcripts and historical logs.
  • Identifying and managing stalled agents that have been inactive for over 30 minutes to optimize compute resources.
  • Tracking total token consumption and task progress across a distributed Openclaw Skills environment.

How Subagent Dashboard Works

  1. The dashboard launches a local Flask server that monitors the specific data directories used by Openclaw Skills.
  2. It scans sessions.json and runs.json to identify active, stalled, and recently completed agents based on timestamp data.
  3. The UI performs an auto-refresh every 3 seconds to pull the latest metadata, token counts, and transcript updates from the file system.
  4. Users interact with the web interface to drill down into specific agent logs or identify bottlenecks in the automated workflow.

Subagent Dashboard Setup

To get started with this Openclaw Skills addition, navigate to the scripts directory and run the provided shell script:

cd workspace/skills/subagent-dashboard/scripts
./start_dashboard.sh

Alternatively, you can set up a Python virtual environment and run the dashboard manually:

python3 -m venv venv
source venv/bin/activate
pip install -r ../requirements.txt
python3 dashboard.py

Subagent Dashboard Data Schema & Taxonomy

The dashboard relies on specific file structures managed by the subagent-tracker within the Openclaw Skills framework to populate the UI:

File Path Description
~/.openclaw/agents/main/sessions/sessions.json Master list of all current active sessions and subagent metadata.
~/.openclaw/agents/main/sessions/*.jsonl Individual transcript files containing the raw interaction logs for each agent.
~/.openclaw/agents/main/subagents/runs.json Task mapping data used to correlate runs with specific agent outcomes and statuses.

Sessions are automatically categorized as Main (orchestrator), Subagent, or Cron based on their metadata badges.

Subagent Dashboard Advanced Features

  • Stalled Agent Detection: Automatically flags any agent that has not produced activity for more than 30 minutes, allowing for rapid intervention.
  • Real-Time Transcript Viewing: Access the full interaction history of any running subagent directly from the browser.
  • Historical Activity Tracking: Displays recently completed tasks (within 24 hours) even after the session has been terminated, provided they have a success status.
  • Flexible Port Configuration: Easily change the server port via the PORT environment variable to fit into your existing Openclaw Skills infrastructure.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*