microsaas-factory_skill

Overview

This skill automates building and deploying micro‑SaaS products from a ConvertFlow template. It generates a complete product config, clones and customizes the template, runs a build pipeline, and prepares a Vercel deployment—stopping for human approval before publishing. Triggered from Telegram (e.g., "build name: description") or from selected high‑scoring ideas.

How this skill works

The skill can run in Quick Build mode (generate product_config.json from a name and description) or Build mode (use a provided product_config.json). It clones the template, injects configuration into targeted files, compiles TypeScript (with auto‑fix attempts for build errors), and pauses for user approval before deploying. It reports progress to Telegram and updates a products registry with build status.

When to use it

• Create a new micro‑SaaS quickly from an idea or brief description
• Automate repetitive template cloning and file customization tasks
• Build and test product code on a CI node before manual deployment
• Maintain an archive of all generated products and their statuses
• Deploy only after human review and credential confirmation

Best practices

• Provide a clear product name and concise description when using Quick Build mode
• Review generated product_config.json and reply "go" to continue with build
• Keep sensitive keys out of automation; provide test Stripe keys only when deploying
• Inspect build logs if the first build attempt fails; allow up to three auto‑fix attempts
• Confirm deployment explicitly by replying "deploy" to avoid accidental publishes

Example use cases

• Turn a strong idea into a working ConvertFlow‑based product with one Telegram command
• Convert internal feature experiments into isolated micro‑SaaS demos for user testing
• Clone and customize multiple variants of a core template for A/B testing
• Archive historic product builds and statuses in the central products.json registry
• Use build mode to reproduce or rebuild a previously defined product_config.json