{"ok":true,"kind":"lesson","integrity":{"knowledge":"free","capability":"paid_x402_optional","rule":"Every concept on a path is fully teachable without purchase. Skills are optional implementations, never required reading.","display":"Learn free. Buy only if you want a ready-to-run sealed pack."},"lesson":{"schema_version":"lvl-knowledge-lesson-v1","id":"path-ma-01-why-workflows-not-one-prompts","category":"path","path_ids":["deploy-multi-agent-workflow"],"title":"Workflows beat one giant prompt","summary":"Multi-step agent systems need ordered stages, handoffs, and failure isolation — not a single mega-prompt.","body":"## The trap\n\nStuffing every instruction into one prompt feels simple until a mid-step fails and the whole run is garbage.\n\n## Workflow thinking\n\n1. **Stages** with clear inputs/outputs  \n2. **Handoffs** that pass structured artifacts  \n3. **Gates** before irreversible actions  \n4. **Observability** per stage  \n\n## Marketplace angle\n\nSkills map cleanly to stages. Bundles package a known-good order. Compatibility graphs answer “what else do I need?”\n\n## Takeaway\n\nCompose small, testable capabilities. Order them. Gate them.","concepts":["multi-agent","workflow","handoff"],"related_skills_optional":[],"sort_order":10,"estimated_minutes":7,"last_updated":"2026-07-22T12:00:00.000Z","freshness":{"last_updated_at":"2026-07-22T12:00:00.000Z","days_since_update":0,"band":"fresh","score_0_100":95,"label":"Fresh","display":"Fresh · updated 2026-07-22 (0d ago)"},"integrity":{"knowledge_free":true,"skills_optional":true,"note":"This lesson is complete without purchasing any skill."},"related_bundles_optional":[],"url":"https://lvlltd.com/learn/lessons/path-ma-01-why-workflows-not-one-prompts/","updated_at":"2026-07-22T21:10:40.587Z"},"note":"Knowledge free. related_skills_optional are implementations, not requirements."}