nfagent

Shipped private

Email generation workflow using multiple AI agents to check brand voice, authenticity, deliverability, and key messaging.

Repository

1 commits over 1 day

What it is

An agent-based system that generates and validates marketing emails through a multi-stage workflow. The system orchestrates specialized agents (authenticity checker, author, brand voice checker, deliverability checker, and key message/CTA checker) to produce polished email content. Built using the Mastra framework for agentic workflows, it represents an experiment in coordinated AI-driven content creation.

Why I built it

Likely built to explore multi-agent collaboration patterns in content generation, specifically testing how specialized agents can work together to ensure email quality across multiple dimensions (authenticity, brand consistency, deliverability, messaging). The "NarrativeFlow Agents" name and PRD presence suggest this was a learning project for agentic software engineering principles.

Tech stack

TypeScript, Mastra framework (for agent orchestration and workflows). Includes Postman collection for API testing and unit tests for workflow validation.

Timeline & effort

Last active Jul 9, 2025 — 1 commit across 1 active day (spanning 0 days).

Minimal effort: a single commit on one day (July 9, 2025). This appears to be an initial scaffold or proof-of-concept pushed in one session.

Did it turn out useful?

Single commit with no follow-up activity suggests this was a one-time exploration rather than a developed tool. The presence of PRDs (including an "improvements.md") and test scaffolding indicates planning for future work, but the project never progressed beyond the initial setup. No README or deployment artifacts confirm it remained an early experiment.

#agents #email-generation #mastra #content-validation #workflow-orchestration #typescript