💡 AI Pipeline Templates & Preset Library

[Weegy]

Overview

Every Payload user who wants a workflow writes code. Every Strapi user who wants automation hires a developer. Numen has the pipeline — but right now, every user starts from scratch. Pipeline Templates eliminate that friction.

The Problem

• Payload CMS: workflows require TypeScript config files
• Strapi: no workflow system at all
• Directus: AI is field-level, no pipeline concept
• Numen: powerful pipeline, but no saved/shareable configurations

New users face a blank canvas. Power users rebuild the same pipeline configs repeatedly across projects.

Proposed Feature: AI Pipeline Templates & Preset Library

Core Capabilities

Save & Name Pipeline Configurations

• Save current pipeline settings (personas, providers, SEO rules, review criteria) as a named template
• Templates stored per-project or globally

Built-In Template Library
Ship with production-ready presets:

• 📝 Tech Blog Post — technical tone, SEO-heavy, code example awareness
• 📣 Product Launch Announcement — marketing persona, urgency, CTA-focused
• 📱 Social Media Pack — short-form, platform-aware (Twitter/LinkedIn/Instagram variants)
• 📧 Newsletter Edition — conversational, subscriber-first, Ghost-competitive
• 🌍 Multilingual Article — i18n-aware, localization pipeline
• 🛒 Product Description — e-commerce optimized, feature-benefit structure

Template Sharing

• Export templates as JSON
• Import from URL or file
• Community template registry (future)

One-Click Instantiation

• Select template → pipeline pre-configured → start with brief
• Template parameters (e.g., target keyword, product name) prompted at runtime

Competitor Context

This is a category-defining feature. No CMS competitor has a pipeline template system because no CMS competitor has an AI pipeline. This ships in a space competitors literally cannot enter without rebuilding their architecture.

Why Now

Directus is actively building AI tooling (@directus/ai v1.3.0 shipped March 10). They're building field-level AI primitives. We must establish pipeline supremacy before they layer workflow logic on top of their field AI.

Acceptance Criteria

• Template save/load from pipeline configuration UI
• Built-in library ships with ≥5 presets
• Template JSON export/import
• Template parameterization (runtime variable injection)
• Template visible in pipeline monitor as "running template: X"