MktClaw for Openclaw

MktClaw is a multi-agent AI marketing agency skill that orchestrates 10 specialized roles to deliver production-ready campaign strategies, live streaming scripts, and public relations plans.

qomob
v1.0.1
Jun 15, 2026
0
716
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install mktclaw

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install mktclaw using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is MktClaw?

MktClaw is a highly optimized agency skill designed to run within the Openclaw Skills ecosystem. Instead of merely offering general advice, MktClaw acts as a fully-featured, demand-to-delivery marketing team that can assume 10 distinct professional roles. It translates raw user requirements (briefs) directly into professional-grade assets, including full-funnel campaign tactics, brand positioning reports, media schedules, and crisis communications.

By utilizing an advanced three-layer context loading architecture (consisting of Hot, Warm, and Cold layers), this addition to the catalog of Openclaw Skills ensures rapid execution and highly accurate outcomes. It prevents token bloat by dynamically loading role-specific instructions and industry-specific benchmarking data only when strictly required, striking a perfect balance between efficiency and domain expertise.

MktClaw Use Cases

  • Drafting end-to-end e-commerce campaigns featuring high-converting social media matrices and schedules.
  • Generating structured live streaming scripts with product sequence planning and traffic-attraction strategies.
  • Formulating targeted KOL and influencer collaboration matrix plans based on pre-calculated benchmark databases.
  • Handling urgent PR incidents by providing risk assessments, response playbooks, and official statements.
  • Running Media Mix Modeling (MMM) and attribution analysis to optimize multi-channel marketing budgets.

How MktClaw Works

  1. Intake Stage (Hot Layer): The intake-agent captures and clarifies user requirements through conversational, multi-choice questions to map out budget, timeline, and goals.
  2. Smart Routing (Hot Layer): The router-agent processes the brief and automatically assigns one or more of the 10 available specialized agency personas.
  3. On-Demand Loading (Warm Layer): The workflow loads the designated role-specific main-agent and its associated workflow schema into active memory.
  4. Execution & Delivery: The assigned agent references the Cold Layer (industry frameworks, playbooks, benchmarks) via the Read tool to output the structured deliverables.
  5. Quality Protocol: If multiple solutions or directions are proposed, an automated Tournament and Adversarial self-check system runs an LLM-as-Judge review to filter out sub-optimal creative outputs before delivering the final assets.

MktClaw Setup

To deploy and configure the MktClaw skill in your AI environment, configure your layout to target the correct skill-base paths.

Ensure your agent system has the Read tool configured, and then initialize the directory:

# Clone or move the skill into your local Openclaw Skills runtime directory
mkdir -p skills/mktclaw
cp -r ./* skills/mktclaw/

# Validate config and run workflows
openclaw skills register skills/mktclaw/config.yaml

Configure your environment to permanently context-load the files defined in the Hot Layer (SKILL.md, agents/intake-agent.md, agents/router-agent.md, and references/agency-types.md).

MktClaw Data Schema & Taxonomy

MktClaw organizes its knowledge base, logic, and output constraints into three clear operational layers:

Layer Location Contents & Purpose
Hot Layer / & /agents/ Constantly active. Contains SKILL.md, intake-agent.md, router-agent.md, and transition-agent.md for orchestrating multi-agent tasks.
Warm Layer /agents/types/{type}/ Context-loaded on-demand. Includes role-specific instructions (e.g., creative-agency/main-agent.md) and local execution workflow.yaml files.
Cold Layer /references/ On-demand dictionary references loaded via file reading tools. Houses shared compliance protocols, platform playbooks, and industry-specific databases (SaaS, FMCG, Beauty, etc.).

MktClaw Advanced Features

  • ** Layered Context Hydration:** Restricts context window overhead by keeping specialized frameworks in a dormant Cold Layer, loading them strictly on-demand.
  • Dual Operating Modes: Features a Standard Mode for deep collaborative agency planning and a Fast Mode (One-Click Delivery) for experienced users requiring immediate execution.
  • Adversarial Self-Checking: Leverages Swiss-system tournaments and adversarial agent reviews to grade creative and strategic copy before presentation.
  • Three-Tier Compliance Engine: Validates all generated content against advertising laws, platform guidelines (e.g., Red/Xiaohongshu rules), and brand-safety parameters.
  • Dynamic Industry Customization: Automatically injects vertical KPIs and audience benchmarks for 10 key industries, including Beauty, Food & Beverage, and Consumer Tech.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*