Your AI Revenue GM: Research to Revenue Pipeline

You are checking dashboards your agent should be checking.

You sell on Gumroad. Or Stripe. Or LemonSqueezy. Maybe all three. Every morning you log in, check sales, update a spreadsheet, think about writing that blog post, and then get pulled into actual work. Revenue tracking falls through the cracks. Content never gets written.

Best for solo operators and small teams selling digital products.

What Your AI Revenue GM does: One config file for your team. One config file for your revenue sources. One install command. 3 AI agents tracking money, researching opportunities, and writing content that sounds human.

Your team:

• Raven (Revenue GM): Tracks sales across 5 platforms, manages the content pipeline, delegates to Scout and Quill, runs on a 2-hour heartbeat
• Scout (Research): Competitive analysis, keyword research, market gaps, structured briefs with sources
• Quill (Writer): Turns briefs into drafts that pass AI detection. 80+ AI patterns banned. 10 human signals required per draft.

You check your phone in the morning. Raven collected overnight sales from Stripe and Gumroad. Scout finished a competitor brief. Quill drafted an article from it. Nothing is waiting on you.

---

How it works:

1. Edit team-config.env (your name, role, business, goals)
2. Edit revenue-config.env (your API keys for each platform you use)
3. Run bash install.sh (auto-detects your setup, fills all 3 agents)
4. Start a new chat (Raven is ready)

Raven is live in under 5 minutes. Full pipeline running in 10.

---

What you get (56 files):

your-ai-revenue-gm/
├── team-config.env              <- EDIT THIS (your info, brand voice, goals)
├── revenue-config.env           <- EDIT THIS (your API keys)
├── install.sh                   <- RUN THIS (preflight, auto-detection, --dry-run)
├── README.md                    <- Start here
├── agent-configs/               <- 3 agents, 7 files each
│   ├── raven/                   <- Revenue GM
│   ├── scout/                   <- Research
│   └── quill/                   <- Writer (anti-AI rules built in)
├── skills/                      <- 3 skills
│   ├── revenue-tracker/         <- 5 platform integrations + manual + reports
│   ├── claw-mart-manager/       <- Listing management CLI
│   └── anti-ai-writing/         <- 80+ banned patterns, checklists, reference
├── docs/                        <- Guides and reference
│   ├── CUSTOMIZATION-GUIDE.md
│   └── content-pipeline.md
├── memory-template/             <- Pre-built tracking files
│   ├── pipeline.md
│   ├── revenue-tracker.md
│   ├── topic-queue.md
│   ├── published.md
│   ├── briefs/
│   └── drafts/
└── scripts/                     <- safe-sub.py

Each agent gets: SOUL.md, IDENTITY.md, AGENTS.md, HEARTBEAT.md, USER.md, TOOLS.md, MEMORY.md

---

Revenue tracking: 5 platforms, one config file.

• Stripe — Automated API pull
• Gumroad — Automated API pull
• LemonSqueezy — Automated API pull
• Paddle — Automated API pull
• Thinkific — Automated API pull
• Claw Mart — Listing management (no sales API available)
• Everything else — Manual logging (affiliate, consulting, any source)

All integrations built from official API documentation. Built-in test command verifies each connection with your account before you rely on it.

Daily and weekly revenue reports with burn-rate projections. Data stored as portable JSONL.

---

Anti-AI writing: content that reads human.

Quill ships with a complete anti-detection writing system:

• 80+ banned AI vocabulary patterns (the words detectors flag)
• 10 mandatory human signals per draft (sentence variation, contractions, real examples, direct opinions)
• Pre-delivery self-check before any draft is handed off
• Reference guides for detection tool internals (how GPTZero and Originality.ai actually score)

Your competitors publish AI slop. Your content passes detection.

---

What works without extra setup:

• All 3 agent personalities and configs: no dependencies
• Revenue tracking with manual entries: no API keys needed
• Anti-AI writing rules: built into Quill's personality
• Content pipeline tracking: uses local memory files
• Claw Mart listing management: needs a Claw Mart API key

What this is NOT:

• Not a real-time sales dashboard or notification system
• Not a Claw Mart sales monitor (their API does not support it)
• Not fully autonomous without the heartbeat cron (one-time setup, directions included)

Requirements:

• OpenClaw installed and running
• Python 3 (for safe config substitution)
• bash, curl, jq (for revenue skills)
• Linux, macOS, or Windows (WSL)

---

$19 one-time. Three agents. Five platform integrations. Content that sounds human.

---

Something broken? Tell us. We fix it.

Built by The Agent Crew