home / skills / bdambrosio / cognitive_workbench / osworld-status

osworld-status skill

safe

This skill reports the OSWorld server status, including readiness, uptime, step count, provider and screen config, for quick diagnostics.

npx playbooks add skill bdambrosio/cognitive_workbench --skill osworld-status

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

Files (2)

Skill.md

1.2 KB

---
name: osworld-status
type: python
description: "Get the current status of the OSWorld server. Returns server readiness, uptime, step counter, provider, and screen configuration."
schema_hint:
  value: "ignored"
  out: "$variable"
examples:
  - '{"type":"osworld-status","out":"$status"}'
---

# OSWorld Status Tool (Level 4)

## Input
- No parameters required
- `value` parameter is ignored

## Output
- Note ID (bound to `out` variable) containing:
  - `text`: formatted status information
  - `metadata`: raw status data including:
    - `ready`: boolean - whether environment is initialized
    - `uptime_sec`: server uptime in seconds
    - `step_counter`: number of actions executed
    - `provider`: provider name (e.g., "docker")
    - `headless`: boolean - whether running headless
    - `screen`: dict with width/height

## Configuration
- `OSWORLD_URL` environment variable (defaults to `http://localhost:3002`)
- Or pass `osworld_url` in character config's `osworld_config` section

## Common Workflow
```json
{"type":"osworld-status","out":"$status"}
{"type":"osworld-observe","out":"$obs"}
{"type":"osworld-execute","python":"pyautogui.click(100,200)","out":"$result"}
```

Overview

This skill reports the current status of an OSWorld server so you can quickly check environment readiness and runtime details. It returns human-friendly text plus raw metadata including readiness, uptime, step counter, provider, headless flag, and screen configuration. Use it to verify the server is ready before running observations or automated interactions.

How this skill works

The skill queries the configured OSWorld endpoint (OSWORLD_URL or osworld_config.osworld_url) and reads a JSON status payload. It formats a short status message and exposes the full raw metadata as the note output for downstream steps. No input parameters are required and any provided value parameter is ignored.

When to use it

Before running automated observations or actions to confirm environment readiness
As a health check in orchestration scripts or CI pipelines
To log server uptime and step counter for debugging or metrics
When switching providers or deployment targets to validate configuration
To detect headless vs. headed mode and screen resolution before UI interactions

Best practices

Configure OSWORLD_URL or set osworld_config.osworld_url to point to the correct environment
Run this skill at the start of every workflow and after any environment changes
Treat the raw metadata as authoritative for automation branching (ready, headless, screen)
Record the formatted text and metadata into logs for traceability
Combine with osworld-observe and osworld-execute steps to build robust interaction flows

Example use cases

Pre-flight check in a test runner to skip actions if the server is not ready
CI health probe that aborts deployment if uptime or provider mismatches expected values
Automation flow that adapts clicks based on reported screen width/height
Debugging session that records step_counter to correlate actions with server state
Monitoring script that polls status and triggers alerts on unexpected provider or headless changes

FAQ

Do I need to pass any parameters to get status?

No. The skill requires no parameters; the value parameter is ignored.

How does it locate the OSWorld server?

It uses the OSWORLD_URL environment variable by default or osworld_config.osworld_url if provided in character config.