cj-wang-sowork

openclaw-workspace-sowork

Production-ready OpenClaw workspace template for marketing teams. Includes brand positioning, web research, and content writing skills, plus a full AGENTS/SOUL/TOOLS/USER/IDENTITY/HEARTBEAT/MEMORY workspace system. Battle-tested across 13 markets by SoWork. DeepEval score: 0.940.

cj-wang-sowork 1 1 Updated 1mo ago
GitHub

Install

npx skillscat add cj-wang-sowork/openclaw-workspace-sowork

Install via the SkillsCat registry.

SKILL.md

openclaw-workspace-sowork

A production-ready OpenClaw workspace template for marketing teams — clone it, fill in your brand context, and run.

What This Skill Does

When you install this workspace, your OpenClaw agent gains:

  1. Brand Positioning — Analyze brand USPs, map competitors, generate campaign strategy.
  2. Web Research — Structured market research and trend monitoring across any market.
  3. Content Writing — Brand-aligned content (social, blog, campaigns) for any platform.

Beyond skills, you get a complete workspace file system:

File Purpose
AGENTS.md Boot sequence, routing rules, checklist table
SOUL.md Persona, tone, brand-in-SOUL methodology
TOOLS.md Environment: SSH hosts, API config, TTS voices
USER.md Team context, preferences (main sessions only)
IDENTITY.md Name, emoji, avatar
HEARTBEAT.md Periodic health tasks and content pipeline checks
MEMORY.md Iron-law rules, key learnings (security-gated)

Installation

# Install via npx skills CLI
npx skills add cj-wang-sowork/openclaw-workspace-sowork

# Or clone directly to your OpenClaw workspace
git clone https://github.com/cj-wang-sowork/openclaw-workspace-sowork.git ~/.openclaw/workspace

Quick Usage

After installation, trigger skills with:

@assistant Run brand positioning for [Your Brand]
@assistant Research [topic or market]
@assistant Write [content type] for [platform]

Production Validation

Extracted from real SoWork operations:

  • 13 markets on 1 VM simultaneously
  • ~$50/month total infrastructure cost
  • 3 agents running in parallel (brand, research, content)
  • DeepEval score: 0.940 on brand consistency
  • Token efficiency: Full bootstrap under 150k chars

Multi-Agent Example

See examples/marketing-team/ for a complete 3-agent setup with cost breakdown.

Security

MEMORY.md is security-gated — loaded in main sessions only, never in group chats or sub-agents.

Related