home / skills / openclaw / skills / openfleet

openfleet skill

/skills/serrato1/openfleet

This skill manages your OpenFleet multi-agent workspace from your terminal, enabling task creation, agent management, pulse cycles, and automations.

npx playbooks add skill openclaw/skills --skill openfleet

Review the files below or copy the command above to add this skill to your agents.

Files (2)
SKILL.md
4.6 KB
---
name: openfleet
description: "Manage your OpenFleet multi-agent workspace — create tasks, assign agents, trigger pulse cycles, manage automations, and monitor activity. Full bidirectional control over your autonomous agent fleet."
homepage: https://openfleet.sh
metadata:
  openclaw:
    emoji: "\u25C7"
    requires:
      bins:
        - npx
      env:
        - OPENFLEET_API_KEY
    primaryEnv: OPENFLEET_API_KEY
---

# OpenFleet

Autonomous multi-agent orchestration from your terminal. Create tasks, manage agents, trigger work cycles, and monitor everything — without leaving OpenClaw.

## Setup

1. Get your API key from the [OpenFleet Dashboard](https://app.openfleet.sh) → Developer Settings
2. Set the env var:
   ```
   export OPENFLEET_API_KEY=ofk_your_key_here
   ```
3. Install the skill:
   ```
   clawhub install openfleet
   ```

### Quick verify

Ask your agent: *"List my OpenFleet tasks"*

## 20 Available Tools

### Tasks

| Tool | Description |
|------|-------------|
| `openfleet_list_tasks` | List tasks — filter by status, priority, or free-text search |
| `openfleet_create_task` | Create a task (starts in INBOX, auto-assigned to best agent) |
| `openfleet_get_task` | Get full task details by ID |
| `openfleet_update_task` | Update title, description, status, priority, or tags |
| `openfleet_delete_task` | Archive (soft-delete) a task |
| `openfleet_unblock_task` | Unblock a BLOCKED task with resolution context |
| `openfleet_approve_task` | Approve a REVIEW task → moves to DONE |
| `openfleet_add_comment` | Add a comment visible to agents during execution |

### Agents

| Tool | Description |
|------|-------------|
| `openfleet_list_agents` | List all agents with status, health, and token usage |
| `openfleet_get_agent` | Get full agent details by ID |
| `openfleet_create_agent` | Create a new agent (LEAD, SPECIALIST, or INTERN) |

### Automations

| Tool | Description |
|------|-------------|
| `openfleet_list_automations` | List recurring task automations |
| `openfleet_create_automation` | Create a scheduled automation (hourly → monthly) |
| `openfleet_toggle_automation` | Toggle an automation on/off |
| `openfleet_trigger_automation` | Fire an automation immediately |

### System

| Tool | Description |
|------|-------------|
| `openfleet_trigger_pulse` | Trigger an agent work cycle (health check + assignment + execution) |
| `openfleet_get_workspace` | Get workspace info and configuration |
| `openfleet_parse_input` | Parse natural language into a structured task |
| `openfleet_install_template` | Install a workspace template (e.g. saas-startup, content-pipeline) |
| `openfleet_list_activities` | List recent activity feed entries |

## Usage Examples

### Create a task
```
Create an OpenFleet task: "Build a landing page with hero section, pricing table, and contact form" with HIGH priority and tags frontend, react
```

### Check agent status
```
List my OpenFleet agents and show who is working on what
```

### Trigger a pulse
```
Trigger an OpenFleet pulse to assign queued tasks and start agent work
```

### Manage blocked tasks
```
Show me all BLOCKED tasks in OpenFleet and unblock the one about API keys with the message "Key has been added to environment"
```

### Create a recurring automation
```
Create a daily OpenFleet automation called "Morning standup report" that generates a summary task every morning
```

## MCP Server Details

This skill wraps the `@open-fleet/mcp-server` npm package, which exposes a standard MCP stdio server.

**Manual MCP setup** (if not using ClawHub):
```bash
npx @open-fleet/mcp-server setup
```

The setup wizard auto-detects Claude Code, Cursor, and Windsurf and configures MCP automatically.

**Direct npx invocation** (for custom configs):
```bash
OPENFLEET_API_KEY=ofk_xxx npx -y @open-fleet/mcp-server
```

## OpenClaw + OpenFleet Integration

When you connect OpenFleet with an OpenClaw gateway, this skill completes the **bidirectional link**:

| Direction | What it does |
|-----------|-------------|
| **OpenFleet → OpenClaw** | OpenFleet sends tasks to your gateway for execution |
| **OpenClaw → OpenFleet** | This skill lets OpenClaw manage tasks, agents, and pulse |

### Full setup

1. Start your OpenClaw gateway: `openclaw gateway`
2. Expose it: `cloudflared tunnel --url http://localhost:18789`
3. Connect the tunnel URL + token in [OpenFleet Settings](https://app.openfleet.sh)
4. Install this skill: `clawhub install openfleet`

## Resources

- [OpenFleet Dashboard](https://app.openfleet.sh)
- [API Documentation](https://app.openfleet.sh/docs)
- [GitHub](https://github.com/open-fleet)
- [SDK](https://www.npmjs.com/package/@open-fleet/sdk)

Overview

This skill manages an OpenFleet multi-agent workspace from your terminal or gateway. It provides full bidirectional control: create and update tasks, manage agents and automations, trigger pulse cycles, and monitor activity. Use it to operate an autonomous agent fleet without leaving your development environment.

How this skill works

The skill calls OpenFleet APIs and wraps the @open-fleet/mcp-server for local MCP integration. It exposes tools to list, create, update, and archive tasks; create and inspect agents; manage scheduled automations; trigger agent pulse cycles; and read workspace activity. It also parses natural-language input into structured tasks and can install workspace templates for common pipelines.

When to use it

  • Orchestrating a team of autonomous agents from the terminal or CI
  • Creating, assigning, or approving tasks in an agent-driven workflow
  • Scheduling recurring work via automations (daily reports, batch jobs)
  • Triggering pulses to force assignment and execution of queued tasks
  • Monitoring agent health, token usage, and recent activity logs

Best practices

  • Set OPENFLEET_API_KEY as an environment variable for secure CLI use
  • Tag tasks and set priority to improve auto-assignment and filtering
  • Use parse_input to transform vague requests into structured task payloads
  • Trigger pulse cycles after creating or updating batches of tasks to ensure timely assignment
  • Use automations for repeatable workflows and toggle them off during maintenance

Example use cases

  • Create a high-priority frontend task with tags and let OpenFleet auto-assign the best specialist
  • List BLOCKED tasks, add resolution comment, and unblock to resume work
  • Create a daily automation that generates a morning standup summary task
  • Trigger an immediate pulse to assign queued tasks before a demo or release
  • Install a saas-startup template to bootstrap agents, tasks, and automations for a new project

FAQ

Do I need an API key?

Yes. Set OPENFLEET_API_KEY from the OpenFleet dashboard to authenticate requests.

Can I run this without OpenClaw?

Yes. The skill wraps the MCP server so you can use npx @open-fleet/mcp-server directly for custom setups, but OpenClaw enables bidirectional gateway integration.

How do automations schedule work?

Automations can be created with schedules ranging hourly to monthly and can be toggled on or off or triggered immediately.