aaron-he-zhu/aaron-marketing-skills

120 marketing skills as an AI marketing staff — plugin, portable skills, or an 8-bot team across 7 disciplines (narrative, SEO/GEO, social, email, paid, influencer, launch) on one contract, with 8 auditor gates: TALE · CORE-EEAT · CITE · ECHO · SEND · ROAS · STAR · RAMP.

What it solves

This project provides a comprehensive, installable AI marketing department. It replaces the need for manual prompting by offering a structured library of 120 specialized agent skills across seven marketing disciplines, ensuring that AI-generated marketing work is consistent, high-quality, and follows a professional operational framework rather than relying on "vibes."

How it works

The system is organized as a four-layer "marketing operating system":

  1. Strategy (L1): Focuses on brand narrative using the TALE framework.
  2. Channels (L2): Manages always-on engines including SEO/GEO, Organic Social, Email, Paid Ads, and Influencer marketing.
  3. Orchestration (L3): Handles time-boxed product launches via the RAMP framework.
  4. Protocol (L4): A shared system of record consisting of truth registries and working memory that ensures consistency across all bots.

Skills are written in plain Markdown and follow a strict activation contract (Trigger, Quick Start, Skill Contract, Handoff, Data Sources, Instructions, and Next Best Skill). The system uses small Bash/Python runtimes for deterministic scoring and validation, and it can be deployed as a single plugin, portable skills, or a team of eight specialized bots (including a chief-of-staff bot for routing).

Who it’s for

It is designed for businesses and marketers who want to automate their marketing operations using AI agents without needing to write complex prompts or manage multiple disconnected tools. It is compatible with various agent hosts like Claude Code, Grok Bot, and Hermes.

Highlights

  • 120 Specialized Skills: Covers Narrative, SEO/GEO, Social, Email, Paid Ads, Influencer, and Launch disciplines.
  • Gated Quality: Uses eight specific benchmarks (like TALE, CORE-EEAT, and ROAS) to provide machine-checkable verdicts on content quality.
  • Keyless by Default: Works with user-provided data (Tier 1) without requiring paid API keys as a precondition.
  • AI Staff Mode: Can be deployed as a team of eight named bots where a chief bot routes goals to the appropriate specialist.
  • Human-like Output: Includes an AI-slop detector and banned-phrase list to ensure professional, human-sounding copy.

Related

  • Project
  • Project
  • Project
  • Project
  • Project