Build Mastra Nuxt Pro
Use this skill when the user wants a complete Nuxt + Mastra implementation path with reliability checks.
Trigger
- •"Create Nuxt Mastra agent in production style"
- •"Nuxt + Mastra full setup with checklist"
- •"pnpm Nuxt Mastra pro workflow"
Target Outcome
- •Nuxt frontend chat UX with streaming
- •Mastra plain chat assistant (
chat-agent) instead of weather assistant - •Clear PDCA-style implementation and verification
- •Operational guardrails for env, routing, and regression checks
Stage-Gated Workflow
- •Preflight
- •Scaffold and dependency setup
- •Agent conversion (weather -> chat) with DeepSeek model
model="deepseek/deepseek-chat" - •API route integration
- •UI integration
- •Validation and acceptance
- •Optional hardening
Must-Follow Constraints
- •Use
pnpmonly. - •Keep agent as generic chat assistant unless user asks for tools.
- •Default model for this skill is DeepSeek chat:
model="deepseek/deepseek-chat". - •Do not use OpenAI API as the default model path in this skill flow.
- •Do not include secrets in source/docs/logs.
- •Keep all edits incremental and rollback-friendly.
- •If versions conflict, prioritize installed package docs and types.
Required References
- •
references/quickstart.md - •
references/architecture.md - •
references/troubleshooting.md - •
references/acceptance-checklist.md
Scripts
- •Run
scripts/preflight.shbefore coding. - •Use script output to decide whether to install/fix prerequisites first.
Deliverables
- •change plan + impact surface
- •implementation steps with expected output
- •file-level change list
- •validation commands and pass/fail criteria
- •rollback plan