clawflows

Purpose

This skill automates workflows in OpenClaw by defining multi-step task chains, incorporating conditional logic, triggers, and scheduling to streamline complex operations.

When to Use

Use this skill for repetitive task sequences, like processing data pipelines, conditional decision-making (e.g., if a file exists, then execute), event-driven triggers (e.g., on API call), or scheduled jobs (e.g., daily reports). Apply it in scenarios requiring orchestration, such as integrating multiple OpenClaw tools or external services.

Key Capabilities

• Define task chains: Specify steps in a JSON config, e.g., {"steps": [{"name": "step1", "action": "run_script"}, {"name": "step2", "action": "send_email"}]}
• Conditional logic: Use if-else in configs, e.g., {"condition": "if status == 'success' then next: step2 else: abort"}
• Triggers: Set event-based triggers like webhooks or file changes, e.g., via {"trigger": {"type": "http", "endpoint": "/webhook"}}
• Scheduling: Use cron-style schedules, e.g., {"schedule": "0 0 * * *"} for midnight daily runs
• Error recovery: Built-in retries for failed steps, configurable per step

Usage Patterns

To create and run a workflow, start by defining a config file (JSON/YAML), then use CLI or API to execute. For simple flows, use CLI directly; for programmatic use, integrate via API. Always set environment variables for authentication, e.g., export CLAWFLOWS_API_KEY=$SERVICE_API_KEY. Test flows in isolation before scheduling. Common pattern: Load config, validate, run, and monitor output.