Skip to content

Use Operately with OpenClaw

OpenClaw is an open source AI agent runtime that connects to your tools via skills. With the Operately skill installed, your OpenClaw agent can manage goals, projects, tasks, milestones, and check-ins through natural conversation or autonomous workflows.

Once connected, your OpenClaw agent can:

  • Create and track goals (OKRs)
  • Open projects, add milestones, and manage tasks
  • Post check-ins and status updates
  • Close milestones and projects with retrospectives
  • Query spaces, people, and notifications
  • Run autonomous workflows that combine Operately with other tools

All operations go through the same CLI and API that human users have access to.

To connect Claude, ChatGPT, or another MCP client directly to Operately (without the CLI), see Connect an AI client with MCP.

Terminal window
openclaw skills install operately/operately-cli

This teaches your OpenClaw agent all available Operately commands.

If not already installed:

Terminal window
npm install -g @operately/operately-cli

Create an API token in Operately (Account Settings > API Tokens), then configure it:

Terminal window
operately auth login --token <your-token>

Or set it as an environment variable in your OpenClaw config:

{
"env": {
"OPERATELY_API_TOKEN": "your-token-here"
}
}

Ask your agent to check the connection:

You: "What spaces do I have in Operately?"

The agent will run operately spaces list and show results.

Your OpenClaw agent can post check-ins on your behalf:

You: "Post a check-in on the Website Refresh project.
Status: on track. We shipped the homepage redesign
and started the /vs/clickup page."

The agent runs the appropriate CLI commands to create the check-in with the right project ID and content.

You: "Create a goal called 'Reach 50 self-hosted installs'
in the Growth space. Target: 50, current: 20."
You: "List all tasks in the SEO Sprint project that are
not started yet, then mark the first three as in progress."

OpenClaw agents can also work autonomously via cron jobs or heartbeat checks. For example, an agent can:

  1. Check project milestones every morning
  2. Post a summary of what is due this week
  3. Flag overdue tasks and notify the team
{
"schedule": { "kind": "cron", "expr": "0 9 * * 1-5", "tz": "Europe/Belgrade" },
"payload": {
"kind": "agentTurn",
"message": "Check all active projects in Operately. Post a Monday summary of due milestones and overdue tasks."
}
}
  • Batch operations: The agent can combine multiple Operately commands in one turn (create a project, add milestones, assign tasks).
  • Cross-tool workflows: Combine Operately with other skills (email, calendar, GitHub) for end-to-end automation.
  • Check-in discipline: Set up a recurring cron job to remind you (or auto-post) weekly check-ins.
  • Read-only exploration: The agent can safely query all data without modifying anything. Start there to build confidence.

The Operately CLI exposes 200+ commands across these namespaces:

NamespaceWhat it does
goalsCreate, list, update, close goals and OKRs
projectsManage projects, milestones, check-ins, retrospectives
tasksCreate, list, update, assign tasks
spacesList and manage workspace spaces
documentsCreate and manage docs in resource hubs
peopleQuery team members and profiles
notificationsList and manage notifications
commentsAdd comments to goals, projects, milestones

Run operately help <namespace> for the full command list.