{"id":"5a7017a8-c8d3-4364-b6a3-e25a6f02431a","shortId":"KwK9qv","kind":"skill","title":"pilot-workflow","tagline":"YAML-defined multi-step workflows with orchestration. Use this skill when: 1. You need complex multi-step workflows with conditional logic 2. You want declarative workflow definitions for reuse 3. You need event-driven orchestration across multiple agents Do NOT use this skill ","description":"# pilot-workflow\n\nYAML-defined multi-step workflows with advanced orchestration capabilities. Enables declarative workflow specifications with conditional branching, loops, parallel execution, and event-driven triggers.\n\n## Commands\n\n### Define workflow YAML\n\n```yaml\nname: data-pipeline\nversion: 1.0\n\ntriggers:\n - type: schedule\n cron: \"0 */6 * * *\"\n\nsteps:\n - id: fetch\n agent: tag:api-gateway\n task: \"Fetch data from https://api.example.com/data\"\n\n - id: validate\n depends_on: fetch\n agent: tag:validator\n task: \"Validate data structure and integrity\"\n\n - id: transform\n depends_on: validate\n condition: \"${validate.result.valid} == true\"\n agent: tag:etl\n task: \"Transform data to parquet format\"\n```\n\n### Execute workflow\n\n```bash\n./pilot-workflow-engine.sh workflow.yaml\n```\n\n### Monitor workflow\n\n```bash\npilotctl --json task list --type submitted | \\\n jq -r '.[] | select(.metadata.workflow_id == \"data-pipeline-001\")'\n```\n\n## Workflow Example\n\nComplete workflow engine:\n\n```bash\n#!/bin/bash\n# Pilot workflow engine - execute YAML-defined workflows\n\nWORKFLOW_FILE=$1\nWORKFLOW_NAME=$(yq eval '.name' \"$WORKFLOW_FILE\")\nWORKFLOW_ID=\"${WORKFLOW_NAME}-$(date +%s)\"\nSTEP_COUNT=$(yq eval '.steps | length' \"$WORKFLOW_FILE\")\n\necho \"Workflow: $WORKFLOW_NAME ($STEP_COUNT steps)\"\n\ndeclare -A STEP_RESULTS\ndeclare -A STEP_STATUS\n\n# Execute each step\nfor ((STEP_IDX=0; STEP_IDX