A comprehensive toolkit for bootstrapping, operating, and replaying multi-agent missions in a controlled runtime environment.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install team-pilot
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install team-pilot 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 TeamPilot Multi-Agent Runtime?
TeamPilot is a robust framework designed to manage multi-agent workflows from a clean environment setup to full-scale mission execution. It provides a centralized mission control interface where developers can switch between autonomous execution and manual deterministic modes, making it a vital component of the Openclaw Skills ecosystem.
By facilitating granular task tracking and historical replays, it ensures high-quality outputs across complex AI operations. Whether you are running smoke tests or controlled demos, this skill provides the necessary API and UI scaffolding to observe and influence agent behavior in real-time.
TeamPilot Multi-Agent Runtime Use Cases
- Setting up a multi-agent environment from scratch on a clean machine.
- Running automated smoke tests using autonomous execution modes.
- Conducting deterministic demos with manual task control and progress updates.
- Analyzing agent performance and task progression via the web-based replay interface.
- Integrating agentic workflows into existing pipelines using a specialized API checklist.
How TeamPilot Multi-Agent Runtime Works
- Environment Bootstrap: Initialize the repository and install Node.js dependencies to prepare the runtime.
- Mission Configuration: Define goals and select execution templates within the Mission Control UI.
- Mode Selection: Choose between auto mode for autonomous runs or manual mode for step-by-step control.
- Task Orchestration: Monitor the mission graph as tasks move from agent lanes to result lanes.
- State Management: Real-time updates are pushed via WebSocket and API endpoints for granular progress tracking.
- Analysis and Replay: Completed missions are stored as traces, allowing for scrubbable, frame-by-frame inspection of inputs and outputs.
TeamPilot Multi-Agent Runtime Setup
Ensure you have Git and Node 20+ installed before beginning the setup for these Openclaw Skills.
git clone <TEAM_PILOT_REPO_URL> team-pilot
cd team-pilot
npm install
npm run up
By default, the UI is available at http://localhost:3333. If you encounter a port conflict, you can specify an alternative:
PORT=3334 npm run up
TeamPilot Multi-Agent Runtime Data Schema & Taxonomy
| Entity | Key Fields | Description |
|---|---|---|
| Mission | id, goal, mode |
Defines the primary objective and execution logic (auto/manual). |
| Task | status, progress, output, assignee |
Tracks individual units of work within a mission graph. |
| Run/Trace | missionId, frames |
A historical log of state snapshots for replay and debugging. |
| API State | liveState, ws |
Real-time mission status delivered via WebSocket connection. |
TeamPilot Multi-Agent Runtime Advanced Features
- Manual mode overrides for deterministic control over every task transition.
- Frame-by-frame replay scrubbing to verify agent input/output logic at specific steps.
- WebSocket-based live state monitoring for zero-latency UI updates.
- Standardized content pipelines including Plan, Research, Verify, Build, and Deliver phases.
- Dynamic port configuration to support multiple concurrent runtime instances.
SKILL.md
Loading
Related Openclaw Skills
XClaw Intelligence
A premier social intelligence layer that empowers AI agents to track trending content, analyze influencer profiles, and monitor real-time social dynamics.
XClaw Intelligence
A high-performance intelligence layer providing real-time social data, KOL deep analysis, and trending tweet discovery for AI agents.
task-father
A generator for file-based task state machines that enables durable, long-running workflows using the OpenClaw workspace.
ssh-op
A secure wrapper that dynamically loads SSH private keys from 1Password into an in-memory agent before executing SSH commands.
Excel Spreadsheet Automation
A comprehensive automation skill for reading, writing, analyzing, and visualizing data within Excel and CSV files.
Moltbook
A social networking skill that allows AI agents to post content, reply to threads, and browse trending topics on the Moltbook platform.
Featured*
